Computer-aided diagnosis with a convolutional neural network algorithm for automated detection of urinary tract stones on plain X-ray.

BACKGROUND: Recent increased use of medical images induces further burden of their interpretation for physicians. A plain X-ray is a low-cost examination that has low-dose radiation exposure and high availability, although diagnosing urolithiasis using this method is not always easy. Since the advent of a convolutional neural network via deep learning in the 2000s, computer-aided diagnosis (CAD) has had a great impact on automatic image analysis in the urological field. The objective of our study was to develop a CAD system with deep learning architecture to detect urinary tract stones on a plain X-ray and to evaluate the model's accuracy. METHODS: We collected plain X-ray images of 1017 patients with a radio-opaque upper urinary tract stone. X-ray images (n = 827 and 190) were used as the training and test data, respectively. We used a 17-layer Residual Network as a convolutional neural network architecture for patch-wise training. The training data were repeatedly used until the best model accuracy was achieved within 300 runs. The F score, which is a harmonic mean of the sensitivity and positive predictive value (PPV) and represents the balance of the accuracy, was measured to evaluate the model's accuracy. RESULTS: Using deep learning, we developed a CAD model that needed 110 ms to provide an answer for each X-ray image. The best F score was 0.752, and the sensitivity and PPV were 0.872 and 0.662, respectively. When limited to a proximal ureter stone, the sensitivity and PPV were 0.925 and 0.876, respectively, and they were the lowest at mid-ureter. CONCLUSION: CAD of a plain X-ray may be a promising method to detect radio-opaque urinary tract stones with satisfactory sensitivity although the PPV could still be improved. The CAD model detects urinary tract stones quickly and automatically and has the potential to become a helpful screening modality especially for primary care physicians for diagnosing urolithiasis. Further study using a higher volume of data would improve the diagnostic performance of CAD models to detect urinary tract stones on a plain X-ray.

A case of nontraumatic urethral rupture possibly induced by prostatic abscess.

We herein report a rare case of nontraumatic urethral rupture. The patient presented with oliguria, perineal pain and anorexia, and a 3 cm fistula was located in the perineum. Computed tomography revealed a retroperitoneal abscess invading the urethra, and a retrograde urethrogram revealed the rupture of the membranous urethra. He had no history of perineum trauma or transurethral procedures. Although he underwent a suprapubic cystostomy and conservative therapy by antibiotics, he eventually died.

Predictive ability of renal cortex enhancement in dynamic computed tomography for residual renal function after nephroureterectomy: Comparison with 99m Tc-diethylenetriaminopentacetic acid renography and validation study.

OBJECTIVE: To estimate postoperative residual renal function after radical nephroureterectomy for upper tract urothelial carcinoma using the preoperative dynamic computed tomography renal cortex enhancement ratio in comparison with the split kidney glomerular filtration rate measured by 99m Tc-diethylenetriaminopentacetic acid renography. METHODS: A total of 47 patients who received radical nephroureterectomy and underwent both preoperative dynamic computed tomography and renography were the model-development cohort; and 109 patients who underwent dynamic computed tomography alone were the validation cohort. Postoperative renal function of the unremoved kidney was estimated using the following formulas: preoperative estimated glomerular filtration rate × the percentage of total renal cortex radiodensity for the intact kidney in Hounsfield units obtained from corticomedullary phase images in the computed tomography-based model, or the percentage of the total glomerular filtration rate measured by renography in the nuclear model. The correlation between observed and estimated postoperative renal function was determined. The computed tomography-based prediction model derived from linear regression analysis was validated externally. RESULTS: The correlation of computed tomography-based split renal function with the observed postoperative estimated glomerular filtration rate (r = 0.80) was equivalent to that of nuclear split renal function (r = 0.78). In the validation cohort, the computed tomography-based prediction model showed an equivalently strong correlation (r = 0.78). CONCLUSIONS: The present study showed that the percentage of total renal cortex radiodensity for the intact kidney is a useful tool for predicting unremoved kidney function in upper tract urothelial carcinoma patients, thereby allowing appropriate patient selection for perioperative cisplatin-based combination chemotherapy.

Prognostic Value of Albumin/Globulin Ratio in Patients with Upper Tract Urothelial Carcinoma Patients Treated with Radical Nephroureterectomy.

BACKGROUND/AIM: We investigated the prognostic role of the albumin/globulin ratio (AGR) in patients with upper tract urothelial carcinoma (UTUC) undergoing radical nephroureterectomy (RNU). PATIENTS AND METHODS: AGR was calculated as follows: AGR=serum albumin/(serum total protein-serum albumin). Associations of preoperative AGR with disease-free (DFS) and overall (OS) survival were assessed in 105 patients with UTUC undergoing RNU. RESULTS: Patients with preoperative AGR <1.24 and ≥1.24 were classified into the low (n=46, 44%) and high (n=59, 56%) groups, respectively. The 5-year DFS and OS were 77% and 78%, respectively. On multivariate analysis, high preoperative AGR was an independent predictor for both better DFS (hazard ratio(HR)=0.34, p=0.038) and OS (HR=0.24, p=0.006). The 5-year DFS and OS were significantly longer in the high-AGR group than in the low-AGR group (90% vs. 60%; 89% vs. 65%, both p<0.001). CONCLUSION: The AGR has prognostic value in patients with UTUC undergoing RNU.

Performance of Quick Sequential (Sepsis Related) and Sequential (Sepsis Related) Organ Failure Assessment to Predict Mortality in Patients with Acute Pyelonephritis Associated with Upper Urinary Tract Calculi.

PURPOSE: The Third International Consensus Definitions for Sepsis and Septic Shock Task Force proposed a new definition of sepsis based on the SOFA (Sequential [Sepsis-related] Organ Failure Assessment) score and introduced a novel scoring system, quickSOFA, to screen patients at high risk for sepsis. However, the clinical usefulness of these systems is unclear. Therefore, we investigated predictive performance for mortality in patients with acute pyelonephritis associated with upper urinary tract calculi. MATERIALS AND METHODS: This retrospective study included 141 consecutive patients who were clinically diagnosed with acute pyelonephritis associated with upper urinary tract calculi outside the intensive care unit. We evaluated the performance of the quickSOFA, SOFA and SIRS (systemic inflammatory response syndrome) scores to predict in-hospital mortality and intensive care unit admission using the AUC of the ROC curve, net reclassification, integrated discrimination improvements and decision curve analysis. RESULTS: A total of 11 patients (8%) died in the hospital and 26 (18%) were admitted to the intensive care unit. The AUC of quickSOFA to predict in-hospital mortality and intensive care unit admission was significantly greater than that of SIRS (each p <0.001) and comparable to that of SOFA (p = 0.47 and 0.57, respectively). When incorporated into the baseline model consisting of patient age, gender and the Charlson Comorbidity Index, quickSOFA and SOFA provided a greater change in AUC, and in net classification and integrated discrimination improvements than SIRS for each outcome. Decision curve analyses revealed that the quickSOFA and SOFA incorporated models showed a superior net benefit compared to the SIRS incorporated model for most examined probabilities of the 2 outcomes. The in-hospital mortality rate of patients with a quickSOFA score of 2 or greater and a SOFA score of 7 or greater, which were the optimal cutoffs determined by the Youden index, was 18% and 28%, respectively. CONCLUSIONS: SOFA and quickSOFA are more clinically useful scoring systems than SIRS to predict mortality in patients with acute pyelonephritis associated with upper urinary tract calculi.

A young female case of primary renal carcinoid tumor mimicking fat-poor angiomyolipoma.

Renal mass biopsy is useful for the pathological diagnosis of a small renal mass difficult to distinguish fat-poor angiomyolipoma from renal cell carcinoma radiologically. Here, we report a young female case of a small renal mass suspected as fat-poor angiomyolipoma in imaging studies. The patient received a renal mass biopsy to obtain the correct pathological information. Unexpectedly, the pathological diagnosis was neuroendocrine tumor. She, finally, underwent a right radical nephrectomy as a curative treatment. This case indicates that the radiological findings of primary renal carcinoid tumor can be similar to those of fat-poor angiomyolipoma.

Effect of Preoperative Bacteriuria and Pyuria on Intravesical Recurrence in Patients with Upper Tract Urothelial Carcinoma Undergoing Radical Nephroureterectomy.

BACKGROUND/AIM: We investigated the effect of bacteriuria and pyuria on intravesical recurrence (IVR) in patients with upper tract urothelial carcinoma (UTUC) undergoing radical nephroureterectomy (RNU). PATIENTS AND METHODS: Preoperative bacteriuria and pyuria were defined as urine containing ≥5 bacteria/high-power field (HPF) and >5 white blood cells/HPF, respectively. Their associations with IVR were evaluated in 97 patients with UTUC undergoing RNU. RESULTS: Preoperative bacteriuria [n=15 (15%)] was significantly associated with preoperative pyuria [n=42 (43%), p<0.001]. During follow-up (median of 19 months), 45 (46%) patients developed IVR (median IVR-free survival=38 months). On multivariate analysis, preoperative bacteriuria was an independent predictor for reduced risk of IVR (hazard ratio=0.23, p=0.010). The 2-year IVR-free survival of patients with preoperative bacteriuria and pyuria was significantly longer than that of patients without preoperative bacteriuria (83% vs. 54%, p=0.028) and pyuria (69% vs. 50%, p=0.024), respectively. CONCLUSION: Bacteriuria and pyuria may reduce the risk of IVR in patients with UTUC undergoing RNU.

External validation of TNM-C score in three community hospital cohorts for clear cell renal cell carcinoma.

AIM: To assess the general applicability of TNM-C scoring, which consists of TNM classification and preoperative C-reactive protein concentration, the predictive ability of the TNM-C score was externally validated for patients with clear cell renal cell carcinoma (ccRCC) at three community hospitals. PATIENTS AND METHODS: Seven hundred patients underwent radical or partial nephrectomy after being diagnosed with RCC. Out of the 700 patients, 518 with clear cell carcinoma served as the current study cohort. The predictive ability of the TMN-C score for cancer-specific survival (CSS) was estimated using Harrell's concordance index (c-index). RESULTS: The c-index of the TNM-C score was 0.85 in the entire data set. CSS rates were clearly stratified according to the scoring model (p<0.001). CONCLUSION: Since TNM-C score alone (without pathological details) has a high predictive ability for the prognosis of ccRCC patients, it is generally applicable for use in community hospitals.

Interactions between inducible nitric oxide synthase and cyclooxygenase-2 in response to ischaemia-reperfusion of rabbit bladder.

OBJECTIVE: To investigate the interactions between inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) in response to ischaemia-reperfusion (I/R) of rabbit bladder. MATERIALS AND METHODS: Rabbit bladders were exposed to 2 h of ischaemia by bilaterally clamping the major arteries entering the bladder and then a subsequent 36 h of reperfusion (I/R) with or without intraperitoneal administration of a selective iNOS inhibitor n-(3-(amynomethyl)benzyl)acetamidine (1400W) or a selective COX-2 inhibitor NS-398 given 1 h before killing. The bladder tissues were processed for isometric tension experiments, enzymatic NOS activitiy, tissue contents of nitrite/nitrate (NO(X) ), cyclic guanosine monophosphate (cGMP) and COX activity determined by prostaglandin E(2) (PGE(2) ) production. RESULTS: iNOS and constitutive NOS (cNOS) activities, NO(X) and PGE(2) contents in the bladder tissues at 36 h after reperfusion were significantly higher than those in the sham group with no significant increase in cGMP. Treatment with 1400W abrogated the increases in iNOS activity and NO(X) as well as PGE(2) without changing cNOS activity. In the tension experiments, a NOS substrate, l-arginine, induced detrusor contraction only in the I/R group, which was inhibited by 1400W or NS-398 but not by a selective soluble guanylate cyclase inhibitor 1H-[1,2,4] oxadiazole[4,3-a]quinoxalin-1-one (ODQ). 8-Br-cGMP induced detrusor relaxation in the sham and I/R groups. Also, l-arginine increased NO(X) and PGE(2) in the bladder tissues only in the I/R group, which were inhibited by pretreatment with 1400W. While, l-arginine increased cGMP contents in the I/R group and this increase was suppressed by ODQ but not by 1400W. CONCLUSION: These results show that NO derived from an up-regulation of iNOS after I/R increases COX-2-derived PG via a cGMP-independent mechanism. NO-mediated activation of COX-2 may be an important mechanism for the modulation of bladder function after I/R injury.

Altered nitric oxide synthase, arginase and ornithine decarboxylase activities, and polyamine synthesis in response to ischemia of the rabbit detrusor.

PURPOSE: Little is known about L-arginine catabolism following ischemia in the bladder. We examined the changes in nitric oxide synthase, arginase and ornithine decarboxylase activity, polyamine biosynthesis and the ability to produce nitric oxide following ischemia of the rabbit bladder. MATERIALS AND METHODS: Bladder ischemia was created by ligation of a unilateral bladder artery. At various time points, that is 1, 4, 8, 24, 48 and 72 hours following ligation, the bladders were excised and harvested for determinations. RESULTS: Constitutive nitric oxide synthase, inducible nitric oxide synthase arginase and ornithine decarboxylase activities increased with time, peaking at 48 hours without significant differences between the ligated and nonligated sides in the whole layer. Arginase and ornithine decarboxylase increased mainly in the muscularis following ischemia. Also, putrescine in the muscularis was significantly higher than in the mucosa 48 hours following ischemia. Baseline nitrite/nitrate production in the whole detrusor on the ligated side at 24 hours was significantly lower than that in the normal detrusor. However, nor-hydroxyarginine as an arginase inhibitor and L-arginine increased nitrite/nitrate production in the ischemic detrusor without changing in the normal detrusor. This increasing effect of nor-hydroxyarginine was abolished by nitroarginine methylester as a nitric oxide synthase inhibitor. CONCLUSIONS: Enzymes related to L-arginine catabolism were involved in the early events of ischemic bladder. Arginase may have 2 independent roles, that is 1) activation of arginase/ornithine decarboxylase/polyamines pathways in the muscle injury and remodeling following ischemia, and 2) endogenous negative regulation of nitric oxide production by limiting the L-arginine substrate for nitric oxide synthase.

[Criteria for general practitioners to assess treatment efficacy for urination problems of the elderly].

Criteria for general practitioners to assess treatment efficacy are needed to efficiently treat urination problems in the elderly. Recently, criteria have been developed for urologists to assess the efficacy of benign prostatic hyperplasia treatment using the International Prostate Symptom Score (I-PSS), QoL index and maximal flow rates. We examined whether these criteria can be applied to the general urination problems of the elderly men and women, with 85 male and 16 female subjects aged 50 and over. Furthermore, we set Level I criteria using the I-PSS and QoL indices. Level II criteria using I-PSS, QoL index and post-void residual urine in addition to Level III criteria using I-PSS, QoL index and maximal urine flow rate. Correspondence rates among assessments by these criteria of treatment efficacy were investigated. Ninety-nine of 101 patients (98.0%) were assessed properly using the Level III criteria but the efficacy graded as "fair" and "unchanged" in Level III in each patient should be "poor" and "fair" in reality. Our findings suggest that Level III criteria should be useful for assessing the treatment efficacy of urination problems of elderly men and women besides benign prostatic hyperplasia. All of the patients, whose treatment efficacy was graded as "excellent" or "good", and those with efficacy graded as "unchanged" or "poor", according to the Level I criteria, were appropriately assessed by the Level III criteria. Of 35 patients whose treatment efficacy was graded as "fair" at Level I, the efficacy of 6 patients was graded as "unchanged" according to Level III. All of the patients, whose treatment efficacy was graded as "excellent" or "good", and those with efficacy graded as "poor", according to the Level II, were also appropriately assessed by the Level III criteria. However, the efficacy of 11 and 4 patients of 38 and 35 patients, respectively, whose efficacy was graded as "fair" and "unchanged" at Level II, was graded as "unchanged" and "fair" according to Level III, respectively. We believe that simple criteria including only I-PSS and QoL index are useful for general practitioners to assess treatment efficacy of urination problems in the elderly. It is suggested that patients with efficacy graded as "unchanged", or "poor" should be referred to a urologist but those with efficacy graded as "excellent", "good" and "fair" should be treated continuously. However, 17% of the patients with efficacy graded as "fair" by Level I criteria might be judged "unchanged" from the viewpoint of urologists.

[Initial assessment criteria for urination problems of the elderly--comparison of criteria for patients/caregivers/nurses, general practitioners, and urologists].

As the population ages rapidly, we need to establish a cost-effective system to assess and treat urination problems of the elderly. Recently, criteria have been developed for urologists to assess benign prostatic hyperplasia using the International Prostate Symptom Score (I-PSS), QoL index, urination function (maximal flow rates and residual urine volume) and prostate volume. It is suggested that patients with moderate to severe urination problems as measured by these criteria need to be treated but those with mild problems have no need for treatment. We examined whether these criteria can be applied to the general urination problems of elderly men and women, with 112 men and 21 women aged 50 and over as subjects. Furthermore, we set Level I criteria for the elderly themselves, their families or nurses taking care of them, using I-PSS and QoL index, and Level II criteria for general practitioners, using I-PSS, QoL index and residual urine volume, in addition to Level III criteria for urologists (omitting prostate volume in women). Correspondence rates among severities of urination problems at Level I, II and III, and sensitivities and specificities of Level I and II criteria were investigated. One hundred twenty-one patients of 124 with moderate to severe urination problems at Level III needed to be treated and eight of nine with mild urination problems did not. As a whole, our findings suggested that 129 of 133 patients (97.0%) were correctly assessed at Level III. We found that Level III criteria were valid for general urination problems of the elderly, not just benign prostatic hyperplasia. While 102 and 111 patients were judged as having moderate or severe urination problems, respectively, at Level I and Level II, and all of these were diagnosed as having moderate or severe problems at Level III, 22 (71.0%) of 31 and 13 (59.1%) of 22 patients judged as having mild urination problems at Level I and Level II were diagnosed as having moderate problems at Level III. Although specificities of Level I and Level II criteria were 100%, sensitivities improved from 82.3% (102/124) to 89.5% (111/124), adding residual urine measurement to the I-PSS and QoL scores. Polyuria of more than 2,000 ml per day was noted in 33 patients (24.8%). At Level I, residual urine of 50 ml or more was noted in 9 (29.0%) patients with mild urination problems, in 22 (24.2%) with moderate problems and in 4 (36.4%) with severe problems. Assessment criteria for general practitioners should include a frequency-volume chart and measurement of residual urine. When the elderly are judged as having moderate or severe urination problems according to either Level I or II criteria, they need treatment. However, more than half of the elderly, judged as having mild problems according to those criteria, need treatment from the standpoint of urologists.

[Myelolipoma of the adrenal gland presenting as upper abdominal symptoms: a case report].

A case of adrenal myelolipoma is presented. A 39-year-old woman was admitted to our hospital for further examination of a right adrenal mass, which was found by examination for nausea, vomiting and upper abdominal pain. Imaging analyses by computed tomography and magnetic resonance imaging revealed a round fatty mass. Endocrine study of the adrenal gland showed normal results. Right adrenalectomy was performed. The tumor weighed 240 g and the histological diagnosis was adrenal myelolipoma. Postoperative course was uneventful and upper abdominal symptoms disappeared after surgery.