Components of image intensive diagnosis. Codifying visual features for effusion cytology

Components of visual assessment in the diagnosis of effusions were analyzed using relative operating characteristic. Diagnostic performance in the assessment of malignancy and the specification of metastatic origin was measured for two expert cytologists. The component of performance attributable to feature interpretation was measured in protocols which minimized the effects of clinical information and visual search in the decision process. Feature interpretation, as a process, contributed significantly to the evaluation of malignancy and marginally to the specification of metastatic origin. For each of these diagnostic tasks, the process of feature interpretation was codified in the construction of explicit models. The expert cytologists were asked to define a set of localized visual features that incorporate essential visual elements for diagnosis. These features were evaluated for a set of test cases, and regression models were constructed defining malignancy and metastatic origin. Relative operating characteristic analysis indicated that the predictive value of the models for diagnosis was very similar to the component of human performance attributable to feature interpretation.     

Extramedullary plasma cell tumors: diagnosis by fine-needle aspiration cytology

Nine cases of plasma cell tumors were diagnosed by fine-needle aspiration cytology, the sites of aspiration being soft-tissue swellings in five cases, lymph nodes in three cases, and both soft-tissue swelling and lymph node in one case. The age of the patients ranged from 33 to 74 yrs, with a median of 52 yrs. Only three of these nine cases were clinically diagnosed as multiple myeloma. Unequivocal demonstration of myeloma cells was possible in eight cases. The smear in ninth case showed mostly undifferentiated blast cells with a few cells showing plasmacytoid differentiation. Giant myeloma cells with monstrous segmented nuclei were encountered in two cases. M-component could be demonstrated in serum and/or urine samples in all seven cases in which it was sought.     

Urinary cytology of a canine bladder carcinoma

This is a study of a case of transitional cell carcinoma of the urinary bladder in a dog. Clinical and radiological signs were inconclusive. The morphology of cells exfoliated from the tumour was very similar to that of cells exfoliated from transitional cell carcinomas in human patients. On the basis of this information a diagnosis was made which was confirmed at post-mortem examination. The findings in this case report demonstrate the usefulness of this technique in the diagnosis of poorly differentiated transitional cell carcinoma.     

The value of cytology in the diagnosis of extragonadal germ cell tumors

Extragonadic germinal tumors are frequently mixed tumors. When a metastatic sacrococcygeal teratoma is clinically suspected in children, a yolk-sac tumor component could be judiciously demonstrated by either an elevated serum level of alphafoetoprotein (AFP), or fine needle aspiration cytology. We report the case of a 25 month aged girl presenting a metastatic sacrococcygeal tumor (lymph node and bone metastasis) with high level of AFP (34.100 microg/ml). Fine needle aspiration cytology identified the yolk sac tumor component, characterized by papillae and glandular clusters, composed of pale isomorphic cells with vesicular and nucleolated nucleus and some intracytoplasmic hyaline globules. The yolk sac tumor component could not be identified in the sacrococcygeal surgical specimen, exclusively composed of immature teratoma.     

Aspiration cytology diagnosis of echinococcosis

We report five cases of echinococcosis diagnosed on material obtained by fine-needle aspiration. The sites were the liver in three cases, the thigh in one case, and the retroperitoneum in one case. No complications were observed.     

DNA flow cytometry and bladder irrigation cytology in detection of bladder carcinoma

In this study we assessed the role of DNA flow cytometry (FCM) as an adjunct to bladder irrigation cytology to detect carcinoma of the bladder. We selected only those cases who had urinary symptoms and cystoscopic examination or histology-proven cases of bladder cancer who underwent cystoscopy for a follow-up study. Cystoscopy, cytologic examination, and DNA FCM were performed in every case. There were 9 fresh cases and 21 follow-up cases of proven transitional-cell carcinoma (TCC) of the bladder. Cystoscopy revealed growth in all 9 fresh cases as well as in 11 follow-up cases. Cytology was positive in 16 cases, out of which there were 8 each of fresh and recurrent cases. None of the cases showed positive cytology with negative cystoscopy findings. DNA FCM was positive in 13 cases. Aneuploidy was detected in 5 cases, out of which there were 3 hyperdiploid and 2 hypodiploid cases. Nine cases had high (equal or more than 10%) S and G2-M phase cells, ranging from 10-19.36%. One case showed aneuploidy along with high S-G2M phase. Both cytology and DNA FCM were positive in 9 cases. In 2 cases, DNA FCM showed aneuploidy, but cytology and cystoscopy were negative. The sensitivity and specificity of the bladder wash cytology were 80% and 100%, and those for DNA FCM were 55% and 83.3%, respectively. We conclude that both bladder wash cytology and DNA FCM techniques should be done in all the cases of suspected TCC to detect more number of positive cases.     

Neural network-based prediction of missing key features in vertical GRF-time recordings

Ground reaction force (GRF) time recordings are frequently corrupted due to faulty stepping on a force platform during gait, and this results in the loss of valuable data and the need for additional data acquisition. This paper proposes a new approach based on artificial neural networks for the estimation of lost key gait parameters using the available features of an affected GRF-time trace. GRF-time plots were recorded using a force platform during normal walking of 14 young and 13 elderly individuals. Back-propagation neural network models were developed using features extracted from 466 vertical GRF-time characteristics that would be unaffected as inputs, and the likely affected gait features (e.g. stance time (ST), push-off force, (Fmax2) and push-off time (Tmax2)) as output. Performance of the models in predicting ST, Fmax2 and Tmax2 were tested using data from 30 new gait trials. The neural network model predicted the missing STs with 96.5% (+/-2.6%) accuracy (r>0.9). Accuracy of the ST prediction deteriorated when push-off force/time data were unavailable for the prediction model. Fmax2 and its time were reconstructed with an accuracy of 95.7% (+/-2.8%) and 97.6% (+/-1.7%) respectively. These results suggest that an artificial neural network may be applied to estimate missing ST, Fmax2 or Tmax2 information using features from an affected vertical GRF-time plot, and the method shows good promise for reconstructing gait forces from a corrupted force-time trace.     

Fine-needle aspiration cytology diagnosis of tuberculosis

Fine-needle aspiration cytology of lymph nodes and extranodal swellings in 160 cases showed granulomatous reaction with or without caseation necrosis in 83%. The material was acellular or predominantly composed of necrotic material, polymorphs, and lymphocytes in 17%. The age of the patient ranged from 1.5 to 72 yr. The male to female ratio was 1:1.3. Acid-fast bacilli (AFB) could be demonstrated in 40.6% of cases. In cases associated with cellular reaction and necrosis. AFB positivity was 50.0%, while it was 66.7% in cases with acellular necrotic material.     

Urine cytology of micropapillary carcinoma of the urinary bladder

A case of micropapillary carcinoma (MPC) of urinary bladder is presented, in which the urine smear was studied in detail in an attempt to better characterize the cytologic findings of MPC. When the voided urine was examined in low power, cancer cells were scattered in the specimens as compact papillary/spheroidal clusters composed of pleomorphic cancer cells. Solitary carcinoma cells were occasionally observed. High power view of the smear revealed that the papillae/spheroids consisted of high-grade urothelial carcinoma cells. The cancer cells had pleomorphic nuclei with coarsely granular chromatin and thickened, irregular nuclear membrane, and thick cytoplasm. Histologically, the tumor in the resected bladder appeared as small nests with surrounding hallo both in the luminal surface and in the site of wall involvement. These tightly bound papillary/spheroidal clusters comprised of highly atypical cancer cells were the most specific cytologic finding in the urine of MPC, which were considered as a key diagnostic clue of MPC. The background of the urine smear showed numerous granulocytes and bacilli compatible with cystitis, which is a previously known complication of MPC. Differential diagnoses of MPC from those with pertinent cytologic findings such as conventional UC (including glandular differentiation), and primary/secondary adenocarcinoma of urinary bladder are discussed with a brief review of literature.     

Cytologic diagnosis of adenoid cystic carcinoma in tracheal wash specimens

Two examples of primary adenoid cystic carcinoma of the trachea are described in which the cytologic diagnosis was made from tracheal washings. It is easy to confuse adenoid cystic carcinoma with the more common small-cell neoplasms of the respiratory tract; however, if differential features are kept in mind, such a distinction can be made in the cytologic material. Differential features include cohesive, globular, honeycomb-like arrangements of cells lacking nuclear molding, three-dimensional cell groups with ball-like formations, and clusters of uniform cells with cystic spaces, some containing mucus. These features in both of our cases resulted in unequivocal cytologic diagnoses.     

Urinary cytology for carcinoma of the urinary bladder]

Cytologic examinations of voided urine were reviewed for 82 cases of invasive or noninvasive papillary carcinoma and 25 cases of carcinoma in situ of the urinary bladder. Five of the 82 cases of invasive or noninvasive papillary carcinoma and 3 of the 25 cases of carcinoma in situ were asymptomatic, and malignancy was detected only by urinary cytology. Urinary cytology was positive in 18.5% of the Grade 1, 33.3% of the Grade 2, 75% of the Grade 3 noninvasive carcinomas and in 77.8% of the invasive carcinomas. However, all 25 cases of carcinoma in situ gave positive cytologic results. Malignant cells found in Grade 1 noninvasive papillary carcinoma were slightly atypic in shape. Their nuclei were small but showed mild to moderate hyperchromasia. These malignant cells appeared in small clusters but the number of clusters were few. In Grade 2 or 3 noninvasive papillary carcinoma, malignant cells appeared in small clusters or isolated single cells. Their nuclei were irregular in shape and showed moderate to marked hyperchromasia. In invasive carcinoma and carcinoma in situ, many malignant cells with marked atypia were observed, but in carcinoma in situ, the background of the specimens was clean.     

Microsatellite analysis of urine sediment versus urine cytology for diagnosing transitional cell tumors of the urinary bladder

The aim was to evaluate microsatellite analysis of urine sediment (MAUS) as an alternative method to urine cytology for routine diagnosis of patients with transitional cell tumors (TCT) of the urinary bladder. Urine cytology has the advantage of being non-invasive, fast and cheap, but is of limited value because of its low sensitivity. MAUS has previously been found to be a successful alternative method. However, the experimental set-up of such investigations implied exclusion of samples with unfavorable characteristics and use of a large number of markers. In the present study, MAUS was tested on all samples routinely available and a small panel of markers was selected. The urine sediments of 66 TCT patients and 24 controls were analyzed by MAUS with 16 fluorescent markers and by urine cytology. All samples were analyzed, including the ones of later micturition, with gross hematuria, leukocyturia or absence of visible sediment. In patients with tumors of low grade (grades I-II), MAUS was significantly more sensitive than urine cytology. The two methods were of equivalent diagnostic power in high-grade (grades III-IV), high-stage (pT1-pT4) tumors. A panel of the six most informative markers for MAUS was selected. Although MAUS has an advantage over routine cytology in low-grade, low-stage tumors, an overall sensitivity of 45% is not sufficient for routine clinical use.     

Diagnostic pitfalls of peritoneal washing cytology and the role of cell blocks in their diagnosis

Mesothelial cell hyperplasia, collagen balls, endometriosis, and endosalpingiosis are diagnostic pitfalls on peritoneal washing cytology in women who present with gynecologic lesions. Over an 8-month period, the peritoneal washings from 10 patients undergoing gynecologic surgery for presumed malignancy showed unusual cytologic findings, several of which posed diagnostic difficulties. The washings from four patients with ovarian carcinomas were cellular and contained clusters and strips of cells with cytologic atypia mimicking malignancy. Confirmation of their benign mesothelial origin was confirmed on immunohistochemistry utilizing cell block preparations. In two cases of endometrial endometrioid carcinoma, the washings contained several clusters of cells surrounding and/or admixed with a globular substance. Due to their similarity to endometrial cells, immunohistochemistry was performed on cell block preparations. The cells were positive for cytokeratin and negative for carcinoembryonic antigen and B72.3, confirming their mesothelial origin. In one case, clinically presumed to be a malignant mass, the washings contained tight clusters of cells with mild cytologic atypia admixed with hemosiderin-laden macrophages. In conjunction with the cell block findings, a diagnosis of endometriosis was made. Extensive endometriosis was found on the surgically resected specimen. In two cases, strips of ciliated epithelial cells resembling tubal epithelium were present on the cytologic and cell block preparations, consistent with endosalpingiosis. The peritoneal washings in one case contained several clusters and balls of atypical cells surrounding microcalcifications on cell block preparation. Since calcification within groups of cells in peritoneal washings always raised the possibility of malignancy, a serous carcinoma of the ovary, particularly of borderline malignancy, would have to be excluded. Fortunately, the resected specimen was free of tumor and showed calcified endosalpingiosis on the ovarian surface. Preparation of cell blocks from peritoneal washings is of value in the work-up and management of patients who present with cytologic mimickers of malignancy on fluid cytology.     

Computer assisted image analysis of Feulgen-stained cell nuclei from transitional cell carcinoma of the human urinary bladder.

Computer assisted image analysis was performed in Feulgen-stained cell nuclei from 117 transitional cell carcinomas of the human urinary bladder. The results were compared to corresponding findings from a control group consisting of 27 specimens from non-malignant urothelium. Generally, the mean nuclear area in carcinoma specimens was found to be increased as compared to the control group, both in diploid and non-diploid tumours. The concentration and the distribution of nuclear chromatin, and the relative area and content of non-condensed chromatin were dependent on nuclear DNA-content and nuclear area. These nuclear parameters did not give supplementary information which could be used in discriminating benign urothelial cell nuclei from malignant ones. A multiple regression analysis, performed for diploid and non-diploid tumours separately, did not show any significant correlation between the considered nuclear parameters and clinical stage of the tumours or the patient's survival.