RESUMEN
BACKGROUND AND OBJECTIVE: With emerging evidence to improve prostate cancer (PCa) screening, multiparametric magnetic prostate imaging is becoming an essential noninvasive component of the diagnostic routine. Computer-aided diagnostic (CAD) tools powered by deep learning can help radiologists interpret multiple volumetric images. In this work, our objective was to examine promising methods recently proposed in the multigrade prostate cancer detection task and to suggest practical considerations regarding model training in this context. METHODS: We collected 1647 fine-grained biopsy-confirmed findings, including Gleason scores and prostatitis, to form a training dataset. In our experimental framework for lesion detection, all models utilized 3D nnU-Net architecture that accounts for anisotropy in the MRI data. First, we explore an optimal range of b-values for diffusion-weighted imaging (DWI) modality and its effect on the detection of clinically significant prostate cancer (csPCa) and prostatitis using deep learning, as the optimal range is not yet clearly defined in this domain. Next, we propose a simulated multimodal shift as a data augmentation technique to compensate for the multimodal shift present in the data. Third, we study the effect of incorporating the prostatitis class alongside cancer-related findings at three different granularities of the prostate cancer class (coarse, medium, and fine) and its impact on the detection rate of the target csPCa. Furthermore, ordinal and one-hot encoded (OHE) output formulations were tested. RESULTS: An optimal model configuration with fine class granularity (prostatitis included) and OHE has scored the lesion-wise partial Free-Response Receiver Operating Characteristic (FROC) area under the curve (AUC) of 1.94 (CI 95%: 1.76-2.11) and patient-wise ROC AUC of 0.874 (CI 95%: 0.793-0.938) in the detection of csPCa. Inclusion of the auxiliary prostatitis class has demonstrated a stable relative improvement in specificity at a false positive rate (FPR) of 1.0 per patient, with an increase of 3%, 7%, and 4% for coarse, medium, and fine class granularities. CONCLUSIONS: This paper examines several configurations for model training in the biparametric MRI setup and proposes optimal value ranges. It also shows that the fine-grained class configuration, including prostatitis, is beneficial for detecting csPCa. The ability to detect prostatitis in all low-risk cancer lesions suggests the potential to improve the quality of the early diagnosis of prostate diseases. It also implies an improved interpretability of the results by the radiologist.
Asunto(s)
Neoplasias de la Próstata , Prostatitis , Masculino , Humanos , Prostatitis/diagnóstico por imagen , Prostatitis/patología , Neoplasias de la Próstata/diagnóstico por imagen , Imagen por Resonancia Magnética/métodos , Próstata/diagnóstico por imagen , Imagen de Difusión por Resonancia Magnética/métodos , Estudios RetrospectivosRESUMEN
OBJECTIVES: MR-guided transurethral ultrasound ablation (TULSA) has primarily been investigated for whole-gland prostate ablation, even though the technology is also well-suited for partial gland treatment. The objectives were to perform a clinical service evaluation of partial to whole-gland TULSA for patients with localized prostate cancer (CaP). TULSA was also evaluated as a combined therapy for a subset of patients presenting with both cancer and concurrent benign prostate hyperplasia (BPH). SUBJECTS AND METHODS: This retrospective, consecutive clinical service evaluation included men with histopathologically-confirmed CaP who underwent TULSA either as primary or salvage treatment. The planned ablation was dependent on the individual tumor characteristics, concurrent BPH and patient preferences. The Clavien-Dindo classification was used to record complications. Surgeon-assessed functional outcomes were reported. Early treatment success was defined by negative multiparametric MRI (mpMRI) and lack of prostate specific antigen (PSA) recurrence. RESULTS: Fifty-two consecutive patients (47 treatment-naïve and 5 salvage) were included, with median follow-up of sixteen months and a max of thirty-six months. Baseline median (IQR) age and PSA were 67 years (63-76) and 8.0 ng/ml (5.2-13), respectively. Two Grade IIIa adverse events were observed, with no bowel-related complications. For urinary continence outcomes, 1 patient worsened to 1 pad per day. All patients who were previously potent maintained erectile potency. Of the patient subgroup also seeking treatment for BPH, 83% reported symptom improvement. Median (IQR) PSA nadir after primary treatment was 1.1 ng/ml (0.5-2.1). Early treatment success was 88%. Nine patients underwent a single repeat TULSA. CONCLUSION: Customized prostate ablation with TULSA offers flexible ablation according to patients' disease characteristics and treatment expectations, providing favorable safety and promising early MRI and PSA results. TULSA is a feasible combination therapy for patients with both cancer and concurrent BPH.
Asunto(s)
Imagen por Resonancia Magnética/métodos , Neoplasias de la Próstata/diagnóstico por imagen , Neoplasias de la Próstata/cirugía , Resección Transuretral de la Próstata/métodos , Anciano , Humanos , Masculino , Persona de Mediana Edad , Neoplasias de la Próstata/patología , Estudios Retrospectivos , Resultado del TratamientoRESUMEN
OBJECTIVES: To determine the utility of our transgluteal magnetic resonance imaging (MRI)-guided prostate biopsy approach. PATIENTS AND METHODS: A total of 960 biopsy series, taken within the period of 1 year, were evaluated, including 301 MRI-guided and 659 transrectal ultrasonography (TRUS)-guided biopsies. RESULTS: The positivity rate and proportion of high grade cancers were significantly higher in MRI-guided than in TRUS-guided biopsies. Of 301 MRI-guided biopsies, 65.4% contained cancer while 57.2% of 659 TRUS biopsies contained cancer (P = 0.016). Gleason grade 3 + 3 = 6 disease was observed in 16.8% of 197 MRI-guided and in 36.1% of 377 TRUS-guided biopsies (P < 0.001). There was also a markedly higher quantity of cancer tissue in MRI-guided biopsies. In all cancers, the mean cancer surface area was 64.8 ± 51.6 mm2 in MRI-guided biopsies as compared with 23.0 ± 31.4 mm2 in non-MRI-guided biopsies (P < 0.001). With respect to the tissue quantity, superiority of MRI-guided biopsy was highest in Gleason grade 3 + 3 = 6 cancers (20.9 ± 27.9 vs 5.1 ± 10.2 mm2 ; P < 0.001) and in Gleason grade 3 + 4 = 7 cancers (59.7 ± 38.0 vs 17.7 ± 18.4 mm2 ; P < 0.001). Comparison of biopsy Gleason grades with findings in prostatectomy specimens was possible in 80 patients with MRI-guided and in 170 patients with non-MRI-guided biopsies. This comparison showed a very high but almost identical concordance of TRUS- and MRI-guided biopsies with the prostatectomy specimen findings. With both approaches, undetected high-risk cancers were present in ~10% of patients with low-risk biopsy results. A significant difference was observed, however, in the proportion of patients who had clinically insignificant cancers and who underwent surgery. The proportion of patients with Gleason grade 3 + 3 = 6 carcinoma in their prostatectomy specimen was 11.2% in the post-TRUS biopsy cohort, but only 2.5% in the post-MRI biopsy cohort (P = 0.021). CONCLUSION: MRI-guided transgluteal prostate biopsy has a high detection rate for high-risk carcinomas, while the risk of detecting clinically insignificant carcinomas appears to be reduced. This may by itself lead to a reduction of unnecessary prostatectomies. Overtreatment may be further avoided by better applicability of molecular testing to MRI-guided biopsies because of the excessive amount of tissue available for analysis, especially in patients with potential low-risk carcinomas.