ARG Automated Report Generation Of X-rays
If we go to see a doctor with a bony injury, chest injury, chest pain or shortness of breath, the doctor will typically ask you to get an X-ray. For the X-rays we most frequently need a radiologist to examine the X-ray image and write an interpreting report accordingly. The report helps the doctor t
2025-06-28 16:25:09 - Adil Khan
ARG Automated Report Generation Of X-rays
Project Area of Specialization Artificial IntelligenceProject SummaryIf we go to see a doctor with a bony injury, chest injury, chest pain or shortness of breath, the doctor will typically ask you to get an X-ray. For the X-rays we most frequently need a radiologist to examine the X-ray image and write an interpreting report accordingly. The report helps the doctor to diagnose the problem and recommend the suitable treatment. Specially, nowadays chest X-rays are also helpful in diagnosing COVID-19 infection in lungs because COVID-19 primarily attacks our lungs. Once the X-ray is done the doctor will use the X-ray report prepared by the radiologist and instruct the patient further.
The aim of our project is to make the interpretation of the X-rays easier. Our project will enable the user to interpret the X-ray images without consulting the radiologists. The user will scan the X-ray image and the X-ray report will be automatically generated using AI.
If we go to see a doctor with a bony injury, chest injury, chest pain or shortness of breath, the doctor will typically ask you to get an X-ray. For the X-rays we most frequently need a radiologist to examine the X-ray image and write an interpreting report accordingly. The report helps the doctor to diagnose the problem and recommend the suitable treatment. Specially, nowadays chest X-rays are also helpful in diagnosing COVID-19 infection in lungs because COVID-19 primarily attacks our lungs. Once the X-ray is done the doctor will use the X-ray report prepared by the radiologist and instruct the patient further.
The aim of our project is to make the interpretation of the X-rays easier. Our project will enable the user to interpret the X-ray images without consulting the radiologists. The user will scan the X-ray image and the X-ray report will be automatically generated using AI.
When a patient visits a hospital with a bony injury or chest injury he/she is asked to take an X-ray. Once the X-ray of the particular body part is done, the patient needs to wait for a radiologist to study the image thoroughly and write an interpretation report. The process is very much time consuming and somehow dependent on the availability of the radiologist. Due to the prolonged process sometimes the patient who needs to be treated on urgent basis especially during the current situation of COVID-19 suffers a lot. In order to overcome the problem we are aiming to design an AI based system which can automatically interpret the X-rays and generate the report without involving the human effort.
Project ObjectivesBuild a deep learning model to generate medical report from given x-ray images.
Project Implementation MethodThe project is a mobile application based system to provide remote access to the user.
The user will login to the system and scan the X-ray image, according to the pre-trained AI model the X-ray interpretation report will be generated. The report will be returned to the user as the outcome and will also be saved in the database to maintain the user’s record.
The model architecture breakdown is as follow:
- The training dataset will be gathered and the model will be trained using the same dataset.
- An algorithm will be designed using supervised learning models which will be able to generate an interpretation report of the X-ray films.
- Test data will be provided to the system with true interpretation reports, based on which the accuracy of the model will be tested.
- Scoring of the system results will be performed.
- The model will be integrated with the mobile application.
- The model will be tested over the real-time images of the X-ray films scanned through the phone camera.
- The application will be deployed and will be available for use.
AI has brought many advancements in the field of medical science. Our project aims to be a contribution in the same field. The key benefits of the project includes
- An AI-aided radiology report system that can assist radiologists in this task and reduce variation between reports thus facilitating communication with the medical doctor or clinician.
- A system that can save the precious time of the patient and the doctor at the same time.
- System that can help in defining the urgency for the patient’s treatment.
- Enablement of the user to get the interpretation of the X-ray report just with the use of a mobile application.
- Cost efficient and accessible.
- A system that produces a well-structured, clear, and clinically well-focused radiology report that is essential for high-quality patient diagnosis and care.
- Under the current situation of COVID-19, the system can be much useful to prioritize the patient with severe lung infections and similar conditions.
- The training dataset will be gathered and the model will be trained using the same dataset.
- An algorithm will be designed using supervised learning models which will be able to generate an interpretation report of the X-ray films.
- Test data will be provided to the system with true interpretation reports, based on which the accuracy of the model will be tested.
- Scoring of the system results will be performed.
- The model will be integrated with the mobile application.
- The model will be tested over the real-time images of the X-ray films scanned through the phone camera.
- The application will be deployed and will be available for use.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 78000 | |||
| Google Colab | Equipment | 1 | 4000 | 4000 |
| Firebase ML API | Equipment | 1 | 8000 | 8000 |
| Cloud Database Server | Equipment | 1 | 58000 | 58000 |
| Banners And poster ,report printing etc | Miscellaneous | 1 | 8000 | 8000 |