Insights into Entomopathogenic Nematode Behavior by Using AI Techniques to Advance Sustainable Pest Control Gianluca Manduca1,2,* , Anita Casadei1,3 , Valeria Zeni3 , Giovanni Benelli3 , Cesare Stefanini1,2 and Donato Romano1,2,* 1 The BioRobotics Institute, Sant’Anna School of Advanced Studies, Viale R. Piaggio 34, Pontedera (PI), 56025, Italy 2 Department of Excellence in Robotics and AI, Sant’Anna School of Advanced Studies, Piazza Martiri della Libertà 33, Pisa, 56127, Italy 3 Department of Agriculture, Food and Environment, University of Pisa, Via del Borghetto 80, Pisa, 856124, Italy Abstract Entomopathogenic nematodes (EPNs) are organisms that are often mass-produced as biological control agents (BCAs) to mitigate pesticide-related hazards and foster environmental sustainability. Enhancing our understanding of EPNs biology and their interactions with hosts is crucial for refining the use of EPNs in integrated pest management. This study pioneers an interdisciplinary approach, integrating engineering, and entomology, to investigate the behavior of Steinernema carpocapsae EPN. Employing a novel blend of deep learning and optical flow analysis, a Convolutional Neural Network (CNN) effectively recognizes how nematodes react to host-borne stimuli. Achieving remarkable precision (1) and an overall accuracy of 0.938, the model elucidates EPN behaviors in a prototyped microfluidic platform reproducing a host-environment context. The integration of optical flow analysis highlights an increased motor activity in EPNs when exposed to stimuli, adding novel information on their dynamic responses. This versatile methodology represents a significant advancement in detecting and understanding EPNs responses to diverse stimuli, fostering their use as advanced BCAs in sustainable pest control and environmental management. Keywords deep learning, optical flow, biological control, chemo-ecology, lab-on-a-chip 1. Introduction proposed in our study aims at advancing understanding on biological control strategies and their applications in In biological and ecological research, machine learning integrated pest management programs, aligning with the and lab-on-a-chip technology are offering a unique com- One Health and EcoHealth perspectives [6]. bination, providing deep insights into biological system Microfluidic platforms offer a controlled environment processes [1]. In this multidisciplinary context, our study for studying the organism-environment interactions, pro- focuses on the understanding of the behavioral and mo- viding researchers to mimic and analyze complex biolog- tor displays of Steinernema carpocapsae Weiser (Rhabdi- ical processes [7]. In the field of behavioral research, tida: Steinernematidae), an entomopathogenic nematode many studies utilize methods that integrate microfluidics (EPN) [2] known for its use as biological control agent and computer vision, especially when studying model (BCA) [3]. This species is described in the literature as a species or those with extensive locomotion capabilities nematode exhibiting ambusher behavior [4]. However, [8]. In addition to these established approaches, we have our objective is to evaluate its ability to move and thus introduced the use of Artificial Intelligence (AI) and deep behave as a cruiser nematode when exposed to organic learning to further explore behavioral differences in this molecules. This new understanding could be beneficial context. AI-based techniques have demonstrated promise both for enhancing ethological knowledge of this species in examining motor anomalies in model organisms [9, 10]. and for expanding the range of potential hosts against Similarly, AI methodologies have enabled the utilization which it could be employed as a BCA [5]. The approach of model organisms as biosensors [11, 12]. Our comprehensive study employs a combination of Ital-IA 2024: 4th National Conference on Artificial Intelligence, orga- established microscopic techniques with microfluidics, nized by CINI, May 29-30, 2024, Naples, Italy deep learning, and optical flow analysis, to investigate S. * Corresponding author. carpocapsae behavior and develop an automated work- $ gianluca.manduca@santannapisa.it (G. Manduca); flow. First, infective juvenile stage specimens of S. car- a.casadei2@studenti.unipi.it (A. Casadei); valeria.zeni@phd.unipi.it (V. Zeni); giovanni.benelli@unipi.it (G. Benelli); pocapsae were observed using an inverted microscope to cesare.stefanini@santannapisa.it (C. Stefanini); evaluate their movement. Experiments were carried out donato.romano@santannapisa.it (D. Romano) in a microfluidic environment where they were exposed  0000-0003-0338-441X (G. Manduca); 0000-0002-1499-067X to organic compounds obtained from the feces of Lobe- (V. Zeni); 0000-0001-8971-6010 (G. Benelli); 0000-0003-0989-041X sia botrana. An arena, designed using fast prototyping (C. Stefanini); 0000-0003-4975-3495 (D. Romano) © 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License techniques, aimed for optimal nematode locomotion and Attribution 4.0 International (CC BY 4.0). CEUR ceur-ws.org Workshop ISSN 1613-0073 Proceedings Figure 1: Workflow of the proposed approach. minimal turbulent motion. A deep learning approach and then fabricated by rapid prototyping in a biocom- differentiated between scenarios involving stimulus pres- patible resin (VisiJet® M3 Crystal, 3D Systems). The ence and control conditions, enabling the analysis of microfluidic testing arena, featuring a releasing chamber motor differences attributed to the stimulus presence. (Ø=5mm; h=2mm) and two stimuli chambers (Ø=3mm; Nematodes are utilized as biosensors to gather environ- h=2mm), creates small environments for the analysis of mental information and investigate changes in behavioral nematode behavior without imposing spatial constraints traits in response to stimuli. The integration of optical on their motility, thereby avoiding potential biases. Each flow analysis revealed distinct variations in motor ac- stimuli chamber is connected to the releasing chamber tivity among individuals across different contexts. This through an aisle creating an y-maze arena. The floor of study bridges microfluidics, machine learning, and sus- the test arena is represented by a transparent glass plate tainable integrated pest management, offering insights firmly connected to the base of the upper component by into S. carpocapsae EPN behavior. It contributes to the depositing and curing a polydimethylsiloxane (PDMS) broader context of the convergence of AI, engineering, film (Sylgard 184). To observe and record the behavior of and entomology, advancing strategies for biomedical re- nematodes, an operator introduced the nematodes into search and sustainable pest control. the releasing chamber using a pipette. Subsequently, the miniaturized testing arena was carefully positioned un- der a 3D visual inspection microscope (Nikon TMS) with 2. Methods a total magnification of 25X. This setup allowed for pre- cise and detailed observation of EPN behavior within 2.1. Steinernema carpocapsae the microfluidic system, ensuring accurate recording of We used NEMOPAK SC, a commercial product distributed their responses to the specific stimuli provided in the by BIOPLANET (Cesena, Italy), to obtain and test Stein- chambers. ernema carpocapsae nematodes at infective juvenile stage (IJ). Nematodes were maintained at 4-8°C till the expira- 2.4. Experiment workflow and recordings tion date and solubilized in water at 25°C, with constant stirring and heating, obtained using a magnetic vortex, The nematodes were reactivated in water at a tempera- to keep them viable before trials. ture of 25°C for 20 minutes as arranged by Bioplanet srl. (Cesena, Italy). Meanwhile, the microarena was washed with water and 70% ethanol to remove contaminants, re- 2.2. Tested cue duce surface tension, enhance arena wettability, and im- EPNs are attracted by chemical cues deriving from po- prove nematode locomotion. A second washing was car- tential hosts. Some studies show the attractive effects of ried out using water to eliminate the alcohol traces poten- gut content of hosts on EPN locomotion [13]. Thus, we tially harmful to the EPN. The chip was then filled with used Lobesia botrana 5th instar larvae feces to investigate the medium, 0.9% NaCl physiological solution within responses in S. carpocapsae. which the nematodes move, it is useful to avoid osmotic stress. Solution and nematodes were inserted using a plastic micro pipette. The chip thus assembled is placed 2.3. Microfluidic platform and under the microscope and the stimulus is positioned in experimental apparatus the opposite arenas, alternating localization to avoid bias The miniaturized testing arena was designed in Solid- during the analysis. The videos were recorded under Works (Dassault Systemes, Velizy Villacoublay, France) the microscope for a total of 5 minutes. Video analysis considers only 60 seconds recording to avoid acclima- sured within each subset. Manual labeling of images was tion time and promote next computer vision and deep performed using the online Makesense software. The learning analysis. Videos are recorded by using an RGB dataset also includes approximately 10% background im- camera (48MP, ƒ/1.8) set on the microscope. Following ages without nematodes, distributed in the same 70-20-10 each test, the device is washed and cleaned in accordance ratio. The CNN was trained with a batch size of 8, over with the previously described procedures. 200 epochs, and an image size of 1920x1080 to enhance data quality when dealing with small objects. Afterward, 2.5. Deep learning detection 32 one-minute inference videos were analyzed, evenly split between stimulus and control contexts. Frames were A deep learning approach was utilized to differentiate extracted from each video at a frequency of 1Hz. For each between scenarios involving stimulus presence and con- frame, the CNN detected nematodes distinguished under trol conditions for nematodes. Beyond discriminating stimulus and control conditions. The outcomes were av- between the two contexts, this approach enabled the eraged for each frame to establish the presence or absence analysis of motor activity differences captured through of a stimulus in the scenario. This procedure was iterated images, attributed to the stimulus presence. In this sense, for each video. In evaluating the CNN model, various nematodes have been employed as biosensors to gather metrics such as accuracy, precision, recall, f1-score, and information about the surrounding environment, while mAP were considered. Training and subsequent analyses also investigating changes in behavioral traits in response were performed using Ultralytics in Python with a Tesla to a stimulus using deep learning techniques. To achieve T4 GPU. this objective, a YOLOv8n Convolutional Neural Network (CNN) model pre-trained on the COCO dataset, a widely 2.6. Optical flow analysis used benchmark for object detection tasks [14], was uti- lized. The selection of this CNN model was driven by its For a comprehensive motor activity investigation, an op- versatility and demonstrated effectiveness in object detec- tical flow analysis was integrated into the deep learning tion tasks. YOLO networks have been applied across vari- framework, using the Farnebäck algorithm [19] with spe- ous detection tasks, particularly in the v8 version, encom- cific parameter settings (image scale of 0.5, 5 pyramid passing small object identification in Unmanned Aerial layers, averaging window size of 15, 3 iterations, size of Vehicle (UAV) images [15], evaluating medical face mask the pixel neighborhood used for polynomial expansion adherence in COVID-19 scenarios [16], or also detecting in each pixel of 7, standard deviation of the Gaussian diverse marine species [17]. YOLOv8 employs a CNN used to smooth derivatives for the polynomial expansion structured into three main components: the backbone, of 1.5). A total of 60 videos, balanced between control neck, and head. The backbone, based on a modified CSP- and stimulus contexts, were considered. Given that ne- Darknet53 architecture, facilitates multi-scaled object de- matodes exhibit body movements with a 2Hz frequency tection through a feature pyramid network, generating according to literature [20], frames were sampled at 5Hz five scale features [18]. The neck introduces a PAN-FPN in accordance with the Shannon theorem for the optical architecture, optimizing performance while maintain- flow analysis. First, the optical flow magnitude was com- ing a lightweight design. The detection head consists of puted by comparing two frames. The CNN model was convolutional and fully connected layers for object classi- exploited to obtain the bounding boxes of the detected fication and bounding box regression. Loss functions like nematodes. Then, the mean magnitude value for each BCE Loss and DFL/CIoU are employed. YOLOv8 operates bounding box was considered. Lastly, the maximum mag- as an anchor-free detection model, directly predicting nitude among the averaged values of the bounding boxes object centers. Dynamic sample assignment is achieved was extracted for each frame, to avoid misclassifications through the Task-Aligned Assigner mechanism, further with background. Temporal data were collected, and the improving accuracy and robustness. Architectural en- mean of these values was computed to derive a represen- hancements include module exchanges and convolution tative value for each video. The analysis was conducted replacements, resulting in a model weighing 6.24 MB using OpenCV in Python. in the v8 nano version. The pre-trained CNN model was fine-tuned to identify and distinguish nematodes in 2.7. Statistical analysis the presence of a stimulus from those in a control set- ting. A dataset of 1680 images was obtained from 50 Data were normally distributed (Shapiro-Wilk test, videos, evenly distributed between control and stimulus p>0.01) and homoscedastic (Levene test, p>0.01). There- contexts. The dataset was divided into training (70%), val- fore, statistical significance between stimuli and control idation (20%), and testing (10%) sets, resulting in subsets was established using a t-test. Statistical analyses were comprising 1189, 339, and 152 images, respectively. An performed using JMP Pro 17 software. The threshold was equal distribution of images across both classes was en- set at p=0.05. Figure 2: Deep learning detection results: frame-level predictions for control (A) and stimulus (B) scenarios; along with the video-level normalized confusion matrix (C) and averaged class prediction values (D). Figure 3: Optical flow analysis results: graphical representation of magnitude for a stimulus context sample (A), based on two frames sampled at 5Hz; for two video samples, comparison of temporal profiles (control and stimulus contexts) of optical flow magnitude, evaluating the maximum value within bounding boxes averaged values for each frame (B); comprehensive video-level analysis considering the mean values of the temporal profiles obtained within the video dataset (C). 3. Results identifying the control context, as demonstrated in the normalized confusion matrix in Fig.2(C). Examination of Microscopical observations have allowed us to identify the normalized confusion matrix indicates that the con- the main behaviors performed by S. carpocapsae, includ- trol class achieved 100% accuracy, while the stimulus class ing an increase in movement and the size of oscillations. exhibited 88% accuracy with a 12% error attributed to mis- This clearly demonstrates the presence of an effect due classification as the control class. Fig. 2(D) showcases to tested cue arising from the host on the gestural com- predictions in video analysis. Overall, the video analysis plex of the EPN. Our findings indicate that S. carpocapsae attained an accuracy of 0.938, precision of 1, recall of exhibits not only a sit-and-wait strategy but also demon- 0.875, and f1-score of 0.933 in identifying stimulus pres- strates active host-seeking behavior in response to chem- ence from behavioral traits. The results obtained through ical stimuli within its proximity. The fact that nematodes deep learning validate the observations. The CNN effec- changed their posture and locomotion direction towards tively distinguished between the two contexts (stimulus the stimulus is functional to describe L. botrana feces as presence and control) based on nematodes. Considering an attractive cue. Furthermore, this characteristic has this image-based approach, the differences are attributed been sparsely evaluated in the literature [21] but holds to altered gestures due to the presence of a stimulus. The significant value for enhancing our understanding about subsequent analysis integrating the optical flow aims this species ethology and optimizing its utilization as to investigate changes in motor activity over time be- BCA. The CNN model achieved a precision of 0.632, re- tween the control and stimulus contexts. Fig.3(A) shows call of 0.623, and mAP0.5 of 0.608 on the validation set. a graphical representation of the magnitude for the stim- Predictions for two images outside the training and vali- ulus scenario, obtained through MINMAX normalization. dation processes are shown in Fig.2(A) and (B). Although Initially, the CNN model was utilized to obtain bounding the network exhibits limitations, particularly in individ- boxes for each frame. Subsequently, the mean optical ual identification compared to background, conducting flow magnitude was calculated for each bounding box. video analysis yields promising results with no errors in Finally, the maximum value among the averaged mag- nitude values in the bounding boxes was determined. not to the emitter. Nematode responses were assessed Temporal profiles of this value for the control and stim- using automated motility tracking and computer vision ulus contexts are compared in Fig.3(B) considering two techniques [24]. The proposed integrated methodology video samples. The mean of these magnitude values over significantly enhances our ability to discern and under- time was considered to obtain an overall value for each stand EPN responses to diverse host stimuli. The use video. Fig. 3(C) illustrates the comparison between con- of AI-based techniques and deep learning, integrated trol and stimulus data. Statistical analysis conducted on with microfluidics and optical flow analysis, represents this data demonstrates significance (p=0.010), providing an innovative approach to behavioral studies, offering evidence of observed differences in motor activity. The the possibility of using nematodes as biosensors. Fur- integration of optical flow with deep learning detection ther research on EPN behavior, utilizing pose estimation reveals increased motor activity of nematodes in the pres- and diverse organic stimuli, could enhance commercial ence of the stimulus when compared to the motility of product performance and expand S. carpocapsae’s use control specimens. as a BCA for other phytophagous species, supporting biological control principles [5]. 4. Discussion Acknowledgments The deep learning approach effectively differentiated be- tween scenarios with stimuli and control conditions for This research was supported by the EU H2020 FETOPEN nematodes. Beyond discrimination, it helped analyze Project “Robocoenosis - ROBOts in cooperation with a motor activity the presence of the stimulus. EPNs have bioCOENOSIS” [899520], the PRIN Project “COSMIC - been used as biosensors, gathering environmental data COntrolled Space MIcroecological system supporting and revealing behavioral changes through deep learning eCopoiesis” granted by the Italian Ministry of Educa- techniques. We utilized various methods, including mi- tion, University and Research (MIUR) [2022EY5BXC], and crofluidics, deep learning, and optical flow, to investigate the EU H2020-MSCARISE-2018 “ECOBOTICS.SEA - Bio- these differences and develop an automated workflow inspired Technologies for a Sustainable Marine Ecosys- for studying EPNs. Microfluidics was already employed tem” [824043]. Funders had no role in the study design, in several biological and behavioral assays with different data collection and analysis, decision to publish, or prepa- species [22], but considering nematodes, this technique ration of the manuscript. The authors are grateful to Ms. is common also for the model organism Caenorhabdi- Cristina Piras for her kind support in visualization. tis elegans [23]. Our innovative approach is essential for studying a parasite’s behavior and locomotion in a natural-like condition. AI analysis revealed challenges References at the frame level due to background interference, but [1] J. Zhou, J. Dong, H. Hou, L. Huang, J. Li, High- exploring different architectures may improve results. throughput microfluidic systems accelerated by ar- However, at the video level, the model showed satisfac- tificial intelligence for biomedical applications, Lab tory performance with 100% precision. Optical flow and on a Chip (2024). doi:10.1039/D3LC01012K. statistical analyses highlighted distinctions in motor ac- [2] R. Gaugler, Ecological genetics of ento- tivity between individuals in different contexts. Tests on mopathogenic nematodes, CSIRO, East Melbourne, EPN movement emphasized their ability to sense, move 1993. towards, and enter hosts. AI-based results confirm the [3] W. Hominick, B. Briscoe, F. G. del Pino, J. Heng, preliminary microscopic observations regarding the pres- D. Hunt, E. Kozodoy, Z. Mracek, K. Nguyen, A. Reid, ence of active motility in the tested species and supports S. Spiridonov, et al., Biosystematics of ento- the few assays which consider S. carpocapsae a nema- mopathogenic nematodes: current status, protocols tode with some cruiser’s features [21]. Considering our and definitions, Journal of Helminthology 71 (1997) results it’s clearly observed that S. carpocapsae has an 271–298. doi:10.1017/S0022149X00016096. actual locomotion towards the stimulus, explainable as [4] E. Lewis, R. Gaugler, R. Harrison, Ento- chemotaxis where there is a spatial movement and klino- mopathogenic nematode host finding: response taxis where speciemens change their body conformation to host contact cues by cruise and ambush for- by bending or turning in presence of the organic stimulus. agers, Parasitology 105 (1992) 309–315. doi:10. Nematodes demonstrate aggregate migration; notably, 1017/S0031182000074230. we observed increased presence near the attractor [23]. [5] P. Grewal, V. Converse, R. Georgis, Influence of According to our results, we are able to define larval feces production and bioassay methods on infectivity of cues as a kairomone; in fact, they evoke a physiological two ambush foragers (nematoda: Steinernemati- response in EPNs, which is favorable to the receiver but dae), Journal of Invertebrate Pathology 73 (1999) make5040083. 40–44. doi:10.1006/jipa.1998.4803. [15] Z. Huangfu, S. Li, Lightweight you only look once [6] D. Destoumieux-Garzón, P. Mavingui, G. Boetsch, v8: An upgraded you only look once v8 algorithm J. Boissier, F. Darriet, P. Duboz, C. Fritsch, P. Gi- for small object identification in unmanned aerial raudoux, F. Le Roux, S. Morand, et al., The one vehicle images, Applied Sciences 13 (2023) 12369. health concept: 10 years old and a long road doi:10.3390/app132212369. ahead, Frontiers in Veterinary Science 5 (2018) 14. [16] S. Tamang, B. Sen, A. Pradhan, K. Sharma, V. K. doi:10.3389/fvets.2018.00014. Singh, Enhancing covid-19 safety: Exploring yolov8 [7] C. E. Stanley, G. Grossmann, X. C. i Solvas, A. J. object detection for accurate face mask classifica- deMello, Soil-on-a-chip: microfluidic platforms for tion, International Journal of Intelligent Systems environmental organismal studies, Lab on a Chip and Applications in Engineering 11 (2023) 892–897. 16 (2016) 228–241. doi:10.1039/C5LC01285F. [17] G. Manduca, L. Padovani, E. Carosio, G. Graziani, [8] C. Restif, C. Ibáñez-Ventoso, M. M. Vora, S. Guo, C. Stefanini, D. Romano, Development of an au- D. Metaxas, M. Driscoll, Celest: computer vision tonomous fish-inspired robotic platform for aqua- software for quantitative analysis of c. elegans swim culture inspection and management, in: 2023 IEEE behavior reveals novel features of locomotion, PLoS International Workshop on Metrology for Agri- Computational Biology 10 (2014) e1003702. doi:10. culture and Forestry (MetroAgriFor), IEEE, 2023, 1371/journal.pcbi.1003702. pp. 188–193. doi:10.1109/MetroAgriFor58484. [9] G. Manduca, V. Zeni, S. Moccia, G. Benelli, 2023.10424093. A. Canale, C. Stefanini, D. Romano, Automated [18] J. Redmon, A. Farhadi, Yolov3: An incremental im- image-based analysis unveils acute effects due to provement, arXiv preprint arXiv:1804.02767 (2018). sub-lethal pesticide doses exposure, in: 2023 45th doi:10.48550/arXiv.1804.02767. Annual International Conference of the IEEE Engi- [19] G. Farnebäck, Two-frame motion estimation based neering in Medicine & Biology Society (EMBC), on polynomial expansion, in: Image Analysis: IEEE, 2023, pp. 1–4. doi:10.1109/EMBC40787. 13th Scandinavian Conference, SCIA 2003 Halm- 2023.10340800. stad, Sweden, June 29–July 2, 2003 Proceedings [10] G. Manduca, V. Zeni, S. Moccia, B. A. Milano, 13, Springer, 2003, pp. 363–370. doi:10.1007/ A. Canale, G. Benelli, C. Stefanini, D. Romano, 3-540-45103-X_50. Learning algorithms estimate pose and detect mo- [20] X. N. Shen, J. Sznitman, P. Krajacic, T. Lamitina, tor anomalies in flies exposed to minimal doses of a P. Arratia, Undulatory locomotion of caenorhab- toxicant, iScience 26 (2023). doi:10.1016/j.isci. ditis elegans on wet surfaces, Biophysical Journal 2023.108349. 102 (2012) 2772–2781. doi:10.1016/j.bpj.2012. [11] E. Fazzari, F. Carrara, F. Falchi, C. Stefanini, D. Ro- 05.012. mano, et al., A workflow for developing biohybrid [21] M. J. Wilson, R.-U. Ehlers, I. Glazer, Ento- intelligent sensing systems, in: Proceedings of the mopathogenic nematode foraging strategies–is Italia Intelligenza Artificiale-Thematic Workshops steinernema carpocapsae really an ambush for- co-located with the 3rd CINI National Lab AIIS Con- ager?, Nematology 14 (2012) 389–394. doi:10. ference on Artificial Intelligence (Ital IA 2023), Pisa, 1163/156854111X617428. Italy, volume 3486, 2023, pp. 555–560. [22] A. Nady, A. R. Peimani, G. Zoidl, P. Rezai, A mi- [12] E. Fazzari, F. Carrara, F. Falchi, C. Stefanini, D. Ro- crofluidic device for partial immobilization, chem- mano, Using ai to decode the behavioral responses ical exposure and behavioural screening of ze- of an insect to chemical stimuli: towards machine- brafish larvae, Lab on a Chip 17 (2017) 4048–4058. animal computational technologies, International doi:10.1039/C7LC00786H. Journal of Machine Learning and Cybernetics (2023) [23] J. A. Carr, A. Parashar, R. Gibson, A. P. Robertson, 1–10. doi:10.1007/s13042-023-02009-y. R. J. Martin, S. Pandey, A microfluidic platform [13] P. S. Grewal, R. Gaugler, E. E. Lewis, Host recog- for high-sensitivity, real-time drug screening on c. nition behavior by entomopathogenic nematodes elegans and parasitic nematodes, Lab on a Chip 11 during contact with insect gut contents, The Jour- (2011) 2385–2396. doi:10.1039/C1LC20170K. nal of Parasitology (1993) 495–503. doi:10.2307/ [24] S. D. Buckingham, F. A. Partridge, D. B. Sattelle, 3283373. Automated, high-throughput, motility analysis in [14] J. Terven, D.-M. Córdova-Esparza, J.-A. Romero- caenorhabditis elegans and parasitic nematodes: González, A comprehensive review of yolo ar- Applications in the search for new anthelmintics, chitectures in computer vision: From yolov1 to International Journal for Parasitology: Drugs and yolov8 and yolo-nas, Machine Learning and Knowl- Drug Resistance 4 (2014) 226–232. doi:10.1016/j. edge Extraction 5 (2023) 1680–1716. doi:10.3390/ ijpddr.2014.10.004.