Skip to main content
The William Harvey Research Institute - Faculty of Medicine and Dentistry

Breadcrumb

  1. WHRI
  2. Study
  3. Barts Charity Doctoral Training Programme
  4. Projects
  5. BC-DTP_2026_70

AI for Early Diabetes Detection using Oral and Metabolic Signals

Code: BC-DTP_2026_70

Title: AI for Early Diabetes Detection using Oral and Metabolic Signals

Primary Supervisor: Tuan Pham

Email: tuan.pham@qmul.ac.uk

Institute: Institute of Dentistry

Secondary Supervisor: David Leslie

Email: r.d.g.leslie@qmul.ac.uk

Institute: Institute of Dentistry

Lay Summary:

Type 2 diabetes (T2D) is a major health challenge in East London, where rates of early‑onset diabetes, late diagnosis, and diabetes‑related complications are significantly higher than the national average. Many people develop early disturbances in blood sugar control, known as dysglycaemia, years before diabetes is diagnosed. During this period, it is important to differentiate between early signs (such as subtle changes in HbA1c, fasting glucose, and clinical symptoms) and risk factors (including diet, BMI, hypertension, and cholesterol levels). While early signs may indicate emerging metabolic dysfunction, risk factors are modifiable determinants—if effectively controlled, they can reverse the prediabetic state and restore normal glucose regulation. These rich clinical data are already captured within NHS systems but are rarely analysed in combination or with modern artificial intelligence (AI) methods to support earlier detection and prevention.

This PhD project will use advanced, interpretable AI models to analyse diverse diabetes‑related data and identify patterns that signal early dysglycaemia in multi‑ethnic East London populations, with the ultimate aim of developing a clinical decision‑support tool that could be integrated into NHS systems—such as the NHS app or electronic health records—to alert doctors and patients to increased risk. A novel aspect of the research is the inclusion of oral markers, such as salivary flow, burning mouth or tongue sensations, and subtle changes to the lining of the mouth, which are known to be influenced by metabolic dysfunction but are not currently used in diabetes screening. By combining standard clinical data with oral features, the project will test whether these early oral changes can improve disease prediction and clarify whether oral manifestations precede or are a consequence of diabetes. If successful, this approach could ultimately be extended for national implementation to support earlier detection and prevention of type 2 diabetes.

Aims:

The aim is to create a low‑cost, accurate, and accessible early‑warning system for both clinicians and patients, designed to reflect the real needs of East London’s diverse communities. The system would generate automated alerts within NHS electronic health records and patient‑facing apps, prompting clinicians to review individuals at increased risk and encouraging patients to seek timely advice or lifestyle support. Earlier detection means people can receive treatment and guidance sooner, reducing complications and helping to tackle long‑standing health inequalities.

References:

[1] Balasanthiran A, O'Shea T, Moodambail A, Woodcock T, Poots A, Stacey M, Vijayaraghavan S. Type 2 diabetes in children and young adults in East London: an alarmingly high prevalence. Practical Diabetes 2012;29:193–198a.

[2] Mathur R, Hull SA, Hodgson S, Finer S. Characterisation of type 2 diabetes subgroups and their association with ethnicity and clinical outcomes: a UK real‑world data study using the East London Database. Br J Gen Pract. 2022;72(719):e421–e429.

[3] Abdelhameed F, Giuffrida A, Thorp B, Moorthy MK, Gevers EF. Exploring the surge in paediatric type 2 diabetes in an inner‑city London centre—a decade‑long analysis of incidence, outcomes, and transition. Children (Basel) 2024;11(2):173.

[4] Chowdhury TA, Huda B. A decade of change in diabetes services in a deprived multiethnic inner‑city area: diabetes in Tower Hamlets. Br J Diabetes 2016;16:188–192.

[5] Ethnicity and type 2 diabetes. Diabetes UK. https://www.diabetes.org.uk/about-diabetes/type-2-diabetes/diabetes-ethnicity. Accessed 15 November 2025.

[6] Leslie RD, Ma RCW, Franks PW, Nadeau KJ, Pearson ER, Redondo MJ. Understanding diabetes heterogeneity: key steps towards precision medicine in diabetes. Lancet Diabetes Endocrinol. 2023;11(11):848–860.

[7] Gesteiro E, Megía A, Guadalupe-Grau A, Fernandez-Veledo S, Vendrell J, González-Gross M. Early identification of metabolic syndrome risk: A review of reviews and proposal for defining pre‑metabolic syndrome status. Nutr Metab Cardiovasc Dis. 2021;31(9):2557–2574.

[8] Sagesaka H, Sato Y, Someya Y, Tamura Y, Shimodaira M, Miyakoshi T, Hirabayashi K, Koike H, Yamashita K, Watada H, Aizawa T. Type 2 diabetes: when does it start? J Endocr Soc. 2018;2(5):476–484.

[9] The Lancet Diabetes & Endocrinology. Prediabetes: much more than just a risk factor. Lancet Diabetes Endocrinol. 2025;13(3):165.

[10] Horton WB, Barrett EJ. Microvascular dysfunction in diabetes mellitus and cardiometabolic disease. Endocr Rev. 2021;42(1):29–55.

[11] Zhao X, Chen X, Zhou Z, Zheng J, Lin Y, Zheng Y, Xu R, Hu S, Cui L. Advancements and challenges in salivary metabolomics for early detection and monitoring of systemic diseases. MedComm (2020) 2025;6(10):e70395.

[12] Diabetes and Dry Mouth. Cleveland Clinic. https://my.clevelandclinic.org/health/symptoms/22495-diabetes-and-dry-mouth. Accessed 16 November 2025.

[13] Sánchez Garrido I, Ramírez L, Muñoz Corcuera M, Garrido E, Sánchez L, Martínez Acitores ML, Hernández G, López-Pintor RM. Xerostomia and salivary dysfunction in patients with diabetes mellitus: a cross‑sectional study. J Oral Pathol Med. 2024;53(10):622–636.

[14] Huang S, Zeng X, Deng S, He S, Liu F. Prevalence of xerostomia in patients with type 2 diabetes mellitus: a systematic review and meta‑analysis. BMC Oral Health 2025;25(1):662.

[15] Melo JLMA, Nunes FPES, Coelho CPES, Heller D, Guimarães MDCM, Do T, Borges LGDA, Grisi DC, Dame‑Teixeira N. Low‑level laser and transcutaneous electrical nerve stimulation on salivary glands impact type 2 diabetes mellitus oral microbiome: an interim analysis of a randomized trial. Lasers Med Sci. 2025;40(1):372.

[16] Techatanawat S, Puangmaliwan P, Thanathornwong B, Chairatvit K, Bhattarai BP, Khovidhunkit W, Khovidhunkit SP. Salivary glucose, salivary interleukin‑18, glycemic control, and periodontal status in patients with type 2 diabetes mellitus. Sci Rep. 2025;15(1):38414.

[17] Lu X, Xie Q, Pan X, Zhang R, Zhang X, Peng G, Zhang Y, Shen S, Tong N. Type 2 diabetes mellitus in adults: pathogenesis, prevention and therapy. Signal Transduct Target Ther. 2024;9(1):262.

[18] Samodova D, Stankevic E, Søndergaard MS, Hu N, Ahluwalia TS, Witte DR, Belstrøm D, Lubberding AF, Jagtap PD, Hansen T, Deshmukh AS. Salivary proteomics and metaproteomics identifies distinct molecular and taxonomic signatures of type‑2 diabetes. Microbiome. 2025;13(1):5.

[19] Ebersole JL, Kirakodu SS, Zhang XD, Dawson D 3rd, Miller CS. Salivary features of periodontitis and gingivitis in type 2 diabetes mellitus. Sci Rep. 2024;14(1):30649.

[20] Pham TD, Holmes S, Chatzopoulou D, Coulthard P. Artificial Intelligence in Facial Trauma, Oral Diseases, and Systemic Health. Switzerland: Springer Nature; 2026. 362 p. ISBN: 978‑3‑032‑11530‑0.

Back to top