Predictive Analytics for Tongue Disease Diagnosis: A Comparative Study of Deep Learning Models Suneel Gollapalli, K. Devender, Munjeti Sowjanya, Vadamodula Prasad, R. Aruna, Balajee Maram Proceedings 2025 2nd International Conference on Networks and Soft Computing Icnsoc 2025, 2025 Tongue disease can be the forerunner of health disorder of the system, and correct diagnosis must thus be performed in order to do something at the appropriate time. Predictive analytics are employed within this study for comparison of deep learning architectures - Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and transfer learning architectures viz. VGG16, ResNet50, and EfficientNetB0 - for tongue disease diagnosis. With a dataset of 12,000 tongue images collected over 8 disease classes, models were evaluated on accuracy, precision, recall, and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{F}1$</tex>-score. ResNet50 recorded 94.7 % accuracy over CNN (87.3 %) and RNN (82.1 %). EfficientNetB0 also recorded a 28% inference speedup with near zero performance loss cost. The article illustrates that the power of transfer learning may be utilized for making non-invasive diagnostics less sensitive in order to drive their clinical uptake at a faster speed.
Enhancing the Detection of PCOS using Deep Learning Suneel Gollapalli, P. Kanchanamala, Satish Muppidi, B. Ankitha, Ch. Srikanth, Venubabu Rachapudi 2nd International Conference on Self Sustainable Artificial Intelligence Systems Icssas 2024 Proceedings, 2024 Polycystic Ovary Syndrome (PCOS) is a widespread endocrine disorder affecting women of reproductive age, frequently resulting in infertility due to hormonal abnormalities. Symptoms include severe acne, irregular menstrual cycles, excess body hair, and weight gain. Early and accurate diagnosis of PCOS is critical for effective management and minimizing related health risks. Traditional manual identification of ovarian follicles in ultrasound images by radiologists can lead to misdiagnoses. To overcome this, proposed an innovative PCOS classification approach that integrates the encoder part of a Transformer model with various pretrained CNN models for robust feature extraction from ultrasound images. This hybrid model leverages advanced deep learning and attention mechanisms to significantly improve diagnostic precision. Our study aims to deliver a highly accurate and comprehensive PCOS detection method, enabling early intervention and customized treatment plans. This cutting-edge approach, situated at the intersection of artificial intelligence and women’s health, holds the potential to transform PCOS management and enhance healthcare outcomes, representing a notable advancement in the field.
RECENT SCHOLAR PUBLICATIONS
Enhancing the detection of PCOS using DEEP LEARNING S Gollapalli, P Kanchanamala, S Muppidi, B Ankitha, C Srikanth, ... 2024 2nd International Conference on Self Sustainable Artificial … , 2024 2024 Citations: 11
MOST CITED SCHOLAR PUBLICATIONS
Enhancing the detection of PCOS using DEEP LEARNING S Gollapalli, P Kanchanamala, S Muppidi, B Ankitha, C Srikanth, ... 2024 2nd International Conference on Self Sustainable Artificial … , 2024 2024 Citations: 11
