No abstract

Two key problems in developing a storyboard are (a) the extraction of video key frames and (b) the display of the storyboard. On the basis of our findings from a preliminary study as well as the results of previous studies on the computerized extraction of key frames and human recognition of images and videos, we propose an algorithm for the extraction of key frames and the structural display of a storyboard. In order to evaluate the proposed algorithm, we conducted an experiment, the results of which suggest that participants produce better summaries of the given videos when they view storyboards that are composed of key frames extracted using the proposed algorithmic method. This finding held, regardless of whether the display pattern used was sequential or structural. In contrast, the experimental results suggest that in the case of employing a mechanical method, the use of a structural display pattern yields greater performance in terms of participants' ability to summarize the given videos. Elaborating on our results, we discuss the practical implications of our findings for video summarization and retrieval.

PASH is a common entity that presents with benign imaging features without evidence of subsequent malignant potential. At our institution, in the absence of suspicious features a diagnosis of PASH at core biopsy is considered sufficient, and surgical excision has been obviated.

The surgical management of invasive breast carcinoma differs from that of in situ disease. Invasive carcinoma necessitates axillary lymph node dissection, a procedure that has associated morbidity. We studied 80 cases (66 invasive, 14 in situ) of breast carcinoma that had a histological diagnosis and a preoperative fine-needle aspirate. All slides were reviewed, with 17 cytologic features assessed. We found that six of these features showed a statistically significant difference between the invasive and in situ cases. These were infiltration of fat or stroma by malignant cells (72% of invasive cases demonstrated this feature, but it was not present in any of the in situ cases, P = 0.0002), the presence of myoepithelial cells overlying clusters of tumor cells (seen in 86% of in situ tumors and 7% of invasive cases, P < 0.00001), calcification (present in 71% of in situ and 15% of the invasive group, P = 0.001), foamy macrophages (noted in 64% of in situ tumors and 16% of invasive carcinomas, P = 0.0007), intracytoplasmic vacuoles (seen in 50% of invasive cases and 21% of in situ lesions, P = 0.08), and tubules (present in 30% of invasive and 7% of in situ tumors, P = 0.10). We demonstrate that invasion can be suggested in fine-needle aspirates of carcinomas, provided that true infiltration of fibrofatty connective tissue by neoplastic cells is present. In situ disease has characteristic features, but the presence of invasion cannot be excluded, even in the presence of stromal or adipose tissue fragments without tumor infiltration.

Lobular carcinomas have a distinct natural history with a better response to endocrine therapy and a higher incidence of local recurrence and are more often bilateral. The cytological diagnosis of lobular carcinoma permits a discriminating therapeutic approach with pre-operative Tamoxifen, more generous resection margins, and assessment of the contralateral breast. The cytological features of lobular cancer however are not well defined and the low cell yield from such neoplasms can result in a high false negative rate. To determine whether we could improve the pre-operative diagnosis, we reviewed the cytological features of 112 lobular carcinomas. They had small uniform sized nuclei with irregular outlines and inconspicuous nucleoli. The degree of dissociation was similar to duct carcinomas and the incidence of inadequate aspirates was no higher. We found no features that confidently diagnosed lobular cancer or its sub-types. However, using a combination of features, typing of lobular cancer on aspirated material is possible and should be attempted.

Aims-To assess the relation between the grade and the status of follow up cytology, the completeness of loop excision biopsies with cervical intraepithelial neoplasia (CIN), and the findings at follow up cytology, as well as the diVerences between complete and incomplete exclusion, using the odds ratio. Treatment failure was assessed. Methods-1600 women with CIN (290 CIN1, 304 CIN2, 1006 CIN3) were followed for a minimum of six months and a maximum of 10 years. A database was created and comparisons performed. The mean age of the patients was 37 years. Results-Excision was complete in over 84% of loops. Residual disease and recurrence of high grade dyskaryosis was more common in women with CIN 3 than CIN 2 or 1. No high grade dyskaryosis was seen in the fifth follow up smear in patients with CIN 1 and CIN 2. Residual, recurrent, and persistent disease was most common in patients with incompletely excised CIN at ectocervical and endocervical margins and deep margins of resection than in patients with completely excised CIN. The odds ratios were significantly higher in the women who had incomplete excision of CIN at ectocervical, endocervical, both ecto-and endocervical, and deep margins of resection compared with those with apparent complete excision of CIN lesions. One patient developed invasive squamous cell carcinoma 44 months after loop excision which showed CIN 3 invading endocervical crypts and extending to both ectocervical and endocervical margins of resection. Conclusions-At long term follow up, patients with CIN who have residual disease are at increased risk of persistent disease and should therefore be followed up regularly with cytology and colposcopy. The findings support national policy of returning women with treated CIN of any grade to normal recall after five years except for cases of CIN3 where excision was incomplete or equivocal. In these cases follow up with annual smear for 10 years is recommended. (J Clin Pathol 2000;53:191-196)

Epithelioid gastrointestinal stromal tumors (GISTs) may cause significant diagnostic confusion on fine-needle aspiration (FNA) with carcinomas, neuroendocrine tumors, and melanoma, particularly when metastatic. This study characterizes the cytologic features of nine cases of epithelioid GISTs that were obtained by computerized tomographic guidance in five, by endoscopic ultrasound in three, and from an excised liver tumor in one. Six cases presented as liver masses, one as a perisplenic mass, one as an abdominal mass, and one as a gastric mass. The aspirates revealed mainly single or small clusters of epithelioid cells with a moderate amount of granular to clear cytoplasm, small uniform nuclei with mild to marked nuclear envelope irregularities. Binucleation and intranuclear inclusions were frequent findings. Collagenous stroma was seen in most cases. In three cases, a neuroendocrine tumor was the initial diagnosis. Immunocytochemical staining for c-kit (CD117) was performed on cellblocks in six cases and was positive in five cases. On the subsequent surgical specimen, CD117 was positive in the c-kit-negative cytology case. The diagnosis of GIST should be considered in aspirates of the gastrointestinal tract, liver, mesentery, or abdominal wall mass lesions when epithelioid cells are the predominant cell type. Ancillary studies such as immunohistochemical stains are usually helpful in making a definitive diagnosis.

In Japan, fine needle aspiration biopsy (FNA) of the breast has long been recognized as a useful diagnostic tool, and has been used in many institutions because it provides a rapid, accurate and cost-effective evaluation. However, the use of core needle biopsy (CNB) is increasing, and vacuum assisted biopsy devices have been developed to produce larger specimens for analysis. CNB is useful because the frequency of inadequate specimens is lower than in FNA, and it requires a less invasive procedure than open biopsy. CNB is also more widely used, compared to FNA, because it can provide a more definitive diagnosis of borderline lesions and can be used to distinguish between IDC and ILC. Therefore, the use of CNB with mammographic or ultrasonographic guidance is especially high for non-palpable tumors. FNA is a rapid and non-invasive procedure that is useful for mass lesions. The accuracy of FNA for non-palpable lesions is relatively low, and depends upon the skill of the aspirators, cytoscreeners and cytopathologists involved in the procedure. However, FNA for palpable masses, coupled with a physical and mammographic examination (the so-called triple test) is highly accurate for diagnosis of breast cancer when all three modalities indicate malignancy, and for a benign lesion when all three are negative.