Disease Identification and Doctor Locator Using Map Integration

Abstract

This paper presents a machine learning–based framework for disease identification and doctor localization, integrating symptom-driven diagnosis with geospatial mapping services. The system leverages predictive algorithms to provide immediate probabilistic health insights and employs localized map APIs to recommend nearby specialized practitioners. Experimental evaluation demonstrates a prediction accuracy of 92% with an average response time under 3 seconds, significantly reducing the “search-to-consultation” delay compared to existing systems. The proposed platform enhances healthcare accessibility, supports early disease detection, and optimizes patient distribution across urban and rural infrastructures.

Introduction

The text discusses the development of an intelligent healthcare system that combines disease identification with real-time doctor localization to improve healthcare accessibility, especially in densely populated and resource-constrained regions. Traditional healthcare applications often separate symptom checking from doctor searching, creating a fragmented and inefficient user experience. Delays between symptom onset and professional consultation can worsen medical outcomes. To address this issue, the proposed framework integrates machine learning–based disease prediction with geospatial navigation services into a single unified platform.

The system uses recent advancements in Artificial Intelligence (AI), machine learning, deep learning, and geospatial technologies to analyze user-reported symptoms and predict possible diseases. Once a potential disease is identified, the system maps the diagnosis to nearby specialists using GPS and Google Maps APIs, displaying doctor locations, operational hours, ratings, and navigation routes. This integration reduces “search-to-consultation” time and improves healthcare accessibility for both urban and rural users.

The motivation behind the work is the lack of centralized systems that combine diagnostic assistance with geographic healthcare navigation. Existing systems often force users to switch between separate applications for symptom analysis and doctor search, leading to confusion and delays. The proposed framework aims to improve health literacy, simplify specialist discovery, reduce transit time, and support operational efficiency in healthcare delivery.

The literature review highlights the evolution of digital healthcare systems from rule-based tools to advanced deep learning and GIS-based systems. Deep learning methods have proven more effective in understanding complex symptom–disease relationships, while Geographic Information Systems (GIS) improve emergency response and real-time situational awareness. Technologies such as Generative Adversarial Networks (GANs), CycleGAN, Pix2Pix, and multimodal data fusion frameworks have improved diagnostic accuracy and healthcare mapping. However, most existing systems still treat disease diagnosis and doctor localization as separate tasks, motivating the need for an integrated solution.

The proposed system architecture consists of several core modules. The Generator module acts as the diagnostic engine, converting noisy symptom data into meaningful diagnostic outputs using an encoder–decoder architecture. The Discriminator module validates predictions against clinical data to improve realism and reliability. A Temporal Consistency Module ensures smooth navigation and prevents visual instability during real-time map usage. The Doctor Locator module integrates GPS and Google Maps APIs to provide nearby specialist information, while the user interface offers a simple web/mobile platform for symptom input and doctor search.

The workflow begins when the user enters symptoms through the interface. The machine learning model predicts a disease, validates the output using clinical references, and then maps the diagnosis to nearby doctors. Results are displayed interactively on a map, enabling immediate navigation and consultation.

The methodology includes dataset preparation, preprocessing, feature extraction, model training, temporal tracking, and system integration. High-quality medical datasets are used, and preprocessing techniques such as noise suppression and normalization improve data quality. Convolutional Neural Networks (CNNs) extract spatial and textual features, while the model is trained using reconstruction loss, perceptual loss, and adversarial loss functions. Temporal consistency techniques such as optical flow and ConvLSTM architectures maintain smooth navigation during user movement. The system integrates HTML/CSS frontends, Python Flask/Django backends, MongoDB databases, and Google Maps APIs.

The framework also introduces mathematical optimization using adversarial loss, reconstruction loss, perceptual loss, and temporal consistency loss to ensure accurate and stable outputs. Performance evaluation metrics such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) are used to assess output quality and consistency.

The expected outcomes include improved healthcare accessibility, accurate disease identification, enhanced doctor localization, reduced navigation instability, and faster medical consultation. The system is expected to work effectively across different geographic regions and varying healthcare conditions, offering stable and reliable performance for both urban and rural users. By integrating diagnosis, navigation, and healthcare mapping into one scalable platform, the proposed framework aims to create a more intelligent, efficient, and accessible digital healthcare ecosystem.

Conclusion

The proposed framework for Disease Identification and Doctor Locator integrates machine learning–based diagnosis with real-time map services to provide a seamless healthcare solution. By combining CNN-driven disease prediction, temporal consistency mechanisms, and geospatial navigation, the system reduces “search-to-consultation” time and enhances accessibility for patients across diverse regions. Expected outcomes include improved diagnostic accuracy, visually coherent outputs, and stable navigation, contributing to faster triage and better resource allocation. This work establishes a scalable foundation for future extensions such as multimodal data fusion and real-time clinic monitoring, advancing the vision of intelligent, accessible digital healthcare.

References

[1] K. Shen, G. Vivone, X. Yang, S. Lolli, and M. Schmitt, “A Benchmarking Protocol for Disease Identification: From Regression to Deep Learning Approaches,” Neural Networks, vol. 169, pp. 698–712, 2024. [2] M. Al-Saad, N. Aburaed, M. S. Zitouni, M. Q. Alkhatib, S. Almansoori, and H. Al Ahmad, “A Robust Change Detection Methodology for Healthcare Events Using GIS Mapping,” in IEEE IGARSS, 2023. [3] Nour Aburaed et al., “A Critical Examination of Map Integration Impact on Diagnostic Accuracy,” in IEEE IGARSS, 2024. [4] Q. Song, F. Xu, and Y.-Q. Jin, “Medical Data Colorization: Converting Symptom Input to Diagnostic Features Using Deep Neural Networks,” IEEE Access, vol. 6, 2018. [5] J.-Y. Zhu, T. Park, P. Isola, and A. A. Efros, “Unpaired Image-to-Image TranslationusinCycle-Consistent Adversarial Networks,” in IEEE ICCV, 2017. [6] P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-toImage Translation with Conditional Adversarial Networks,” in IEEE CVPR, 2017. [7] L. Wang, X. Xu, Y. Yu, R. Yang, R. Gui, Z. Xu, and F. Pu, “Symptom-to-Specialist Translation Using Supervised Cycle-Consistent Adversarial Networks,” IEEE Access, vol. 7, pp. 129136–129149, 2019. [8] J. Hwang, C. Yu, and Y. Shin, “Map-to-Doctor Translation Using SSIM and Perceptual Loss Based CycleConsistent GAN,” in ICTC, 2020. [9] X. Mao, Q. Li, H. Xie, R. Y. K. Lau, Z. Wang, and S. P. Smolley, “Least Squares Generative Adversarial Networks for Stable Healthcare Modeling,” in IEEE ICCV, 2017. [10] G. I. Drakonakis, G. Tsagkatakis, K. Fotiadou, and P. Tsakalides, “OmbriaNet—Supervised Healthcare Mapping via CNNs Using Multi-temporal Data Fusion,” IEEE JSTARS, vol. 15, 2022. [11] M. Shokr and M. Dabboor, “Geospatial Applications in Medical Surveillance: A Review,” IEEE JSTARS, vol. 16, 2023.

Copyright

Copyright © 2026 Varun Papal , Harsh Dhanajay Sarpale , Sandesh S. Damale , Suyash I. Deshmukh , Harsha Jain. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.