FindAPhD Weekly PhD Newsletter | JOIN NOW FindAPhD Weekly PhD Newsletter | JOIN NOW

Development of an automated high throughput image scanning system that enables Artificial Intelligence models to sub-classify CT scanning and X-rays at scale

   Department of Mechanical Engineering

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Dr Amir Hajiyavand  No more applications being accepted  Competition Funded PhD Project (Students Worldwide)

About the Project

Lung cancer has the highest mortality rate among all malignant tumours. The prognosis of lung cancer is poor with a 5-year survival rate of 18%.  Rapid and accurate diagnosis of early stage pulmonary nodules plays a crucial role in reducing lung cancer mortality. NHS England have set up Rapid Diagnostic pathways to enable high throughput community-based screening for symptomatic patients. Rapid reporting of large volumes of images will be a fundamental requirement and will inevitably cause significant workload pressures on an already limited human radiology clinical workforce. There are significant advantages of automating the reporting of large image datasets using validated AI models. This will help alleviate bottlenecks in radiology reporting within Rapid Diagnostic pathways. Any learning from this project could easily be applied/transferred to other cancer sites.

The aim of this project is to create an automated digital pathway that enables AI image classification models to rapidly and accurately classify large scale images datasets. 

This project will include: 

• Development and validation the performance of AI models that can accurately analyse DICOM (CT scans and X-ray) files to detect abnormalities.

• Undertake inter-observer variability studies to test the model's performance against human reporting radiologists within a Health-secure Research Environment, specifically looking at model performance, cost and time savings.

• Develop a secure digital framework that can ingest large numbers of DICOM files for rapid classification (normal versus indeterminate versus abnormal) using validated AI models.

• Setting the groundwork to design a prospective AI clinical trial in collaboration with NHS partners that will test the model's performance in the Real World against conventional human expert radiology reporting.

We have developed a cross-disciplinary collaborative team comprising experts from Engineering, Computer Science, Robotics, Medical Physics and Birmingham health partners in order to create a validated AutomateD Diagnostic Rapid Radiology REporting SyStem (ADDRESS). The successful candidate will join the Medical Robotic and Innovation Laboratory (MERIL). This laboratory focuses on research that helps to integrate Artificial Intelligence technologies into healthcare systems. 

All applicants should have a first-class degree in engineering or computer science (or High 2:1) in BEng, MEng or MSc with Distinction and a strong interest in pursuing research in this field. Candidates should be interested in Artificial Intelligence in healthcare and be familiar with computer vision technologies. There would be no requirement to have a medical background for this position. Having previous experience in computer programming, applied artificial intelligence and working with industrial or research would be highly desirable.

Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs