Table 4 Challenges that can be solved using AI and ML techniques across the four cancer types
Challenges that can be solved using AI and ML techniques | Frequency for each tumour type | Total frequency across the 4 tumour types | |||
|---|---|---|---|---|---|
Breast cancer | Lung cancer | Colorectal cancer | Prostate cancer | Across the 4 tumour types | |
Clinical practice related challenges | |||||
 (1) Problems with the accuracy of imaging modalities/tests: there is a need to improve the accuracy of current imaging modalities/tests in terms of reporting and interpretation, rates of false positive and false negative | 22 | 17 | 17 | 11 | 67 |
 (2) Challenges with disease evaluation: improvement in disease evaluation in terms of characterisation, differentiation and staging | 5 | 13 | 10 | 4 | 32 |
 (3) Insufficient standardisation of the care pathway: improvement in the standardisation of the care pathway | 5 | 5 | 4 | 2 | 16 |
 (4) Challenges with disease treatment: improvement in cancer treatment in terms of timing, choices and prognosis | 2 | 4 | 4 | 2 | 12 |
 (5) Challenges with disease recurrence: improve detection/prediction of cancer recurrence | 2 | 2 | 2 | 4 | 10 |
Healthcare professionals related challenges | |||||
 (6) Insufficient expertise | 4 | 4 | 5 | 2 | 15 |
 (7) Human error: elimination of operator dependent error, reduction of interobserver/intraobserver variability | 4 |  |  | 5 | 9 |
Healthcare system related challenges | |||||
 (8) Long waiting lists/times: reduction of long waiting lists/times for diagnosis and treatment | 8 | 11 | 11 | 9 | 39 |
 (9) Lack of resources: optimisation of resources (human, machinery and financial) | 6 | 8 | 8 | 3 | 25 |
 (10) Workload: reduction of workload | 3 |  |  |  | 3 |