Purpose
1) Prostate cancer (Pca) overview
High prevalence: 2nd most common cancer in men [1]
Diagnostic and treatment decision depend on:
Level of Prostate Specific Antigen (PSA) from blood sample
Identification of suspicious lesion using multimodal MRI and scoring from radiologist
(If suspicion) Prostate biopsy to confirm clinically significant cancer (CSC) from uropathologist
2) Prostate cancer characterization:
Histopathologic annotation obtained after biopsy or prostatectomy: Gleason score (GS), from Benign (< 6) to 10 [2]
Lesion with score >= 7: CSC
Reliable measure of cancer aggressiveness
But...
Methods and materials
1) Database
Multi-centric data from various continents (North and South America, Europe, Asia)
Data acquired with different scanners from different manufactures, with a variety of protocols
All patients underwent bi-parametric MRI: high b-value Diffusion-Weighted magnetic resonance Imaging (DWI), ApparentDiffusion Coefficient (ADC), T2-weighted (T2w).
High b-value DWI defined as a DWI with b-value of 2000 if available, otherwise DWI with the closest b-value to 2000 is chosen
Following bp-MRI, some patients underwent in-bore MRI-guided biopsy or radical prostatectomy. Resulting histopathologic analysis were only available for ~24%...
Results
1) Results on validation set:
Detection results:
At 0.5 FP/volume, 87% of significant lesions (GS>6) were detected with model trained on PIRADS and GS, versus 79% for the model trained on GS data only
At 1FP/volume, 91% of significant lesions (GS>6) were detected with model trained on PIRADS and GS, versus 85% at 1FP/volume for the model trained on GS data only.
[Fig 6]
[Fig 7]
Classification results:
AUC = 0.723 for model trained on PIRADS and GS data, versus AUC=0.699 for the model trained...
Conclusion
Our experiments show the interest of combining radiological annotation (PIRADS) with histopathological annotation (Gleason score), in order to improve computer aided detection and diagnosis systems of prostate cancer
Detection results from our model trained on PIRADS and GS annotations is competitive with respect to the best performances found in state of the art:
Saha et al.: 83% of CSC lesions detected at 0.5 FP/volume, 93% at 1FP/volume. Evaluation done on personal evaluation data [9]
Bosma et al.: 84% of CSC lesions detected at 1FP/volume. Evaluation...
Personal information and conflict of interest
N. Debs:
Employee: Guerbet
A. Routier:
Employee: Guerbet
B. Lorenzi:
Employee: Guerbet
L. Wood:
Employee: Guerbet
F. Nicolas:
Employee: Guerbet
M-M. Rohé:
Employee: Guerbet
References
[1] Rawla, P. (2019). Epidemiology of prostate cancer.World journal of oncology,10(2), 63.
[2] Epstein, J. I., Egevad, L., Amin, M. B., Delahunt, B., Srigley, J. R., & Humphrey, P. A. (2016). The 2014 International Society of Urological Pathology (ISUP) consensus conference on Gleason grading of prostatic carcinoma.The American journal of surgical pathology,40(2), 244-252.
[3] Weinreb, J. C., Barentsz, J. O., Choyke, P. L., Cornud, F., Haider, M. A., Macura, K. J., ... & Verma, S. (2016). PI-RADS prostate imaging–reporting and data system: 2015, version 2.European...