Annotated normal CT data of the abdomen for deep learning: Challenges and strategies for implementation

PURPOSE

The purpose of this study was to report procedures developed to annotate abdominal computed tomography (CT) images from subjects without pancreatic disease that will be used as the input for deep convolutional neural networks (DNN) for development of deep learning algorithms for automatic recognition of a normal pancreas.

MATERIALS AND METHODS

Dual-phase contrast-enhanced volumetric CT acquired from 2005 to 2009 from potential kidney donors were retrospectively assessed. Four trained human annotators manually and sequentially annotated 22 structures in each datasets, then expert radiologists confirmed the annotation. For efficient annotation and data management, a commercial software package that supports three-dimensional segmentation was used.

RESULTS

A total of 1150 dual-phase CT datasets from 575 subjects were annotated. There were 229 men and 346 women (mean age: 45±12years; range: 18-79years). The mean intra-observer intra-subject dual-phase CT volume difference of all annotated structures was 4.27mL (7.65%). The deep network prediction for multi-organ segmentation showed high fidelity with 89.4% and 1.29mm in terms of mean Dice similarity coefficients and mean surface distances, respectively.

CONCLUSIONS

A reliable data collection/annotation process for abdominal structures was developed. This process can be used to generate large datasets appropriate for deep learning.

Nuclear morphometry predicts disease-free interval for clinically localized adenocarcinoma of the prostate treated with definitive radiation therapy

Men treated for prostate cancer often have unexpected outcomes despite predictive models based on stage, grade and prostate-specific antigen (PSA). Previous results have indicated that nuclear morphometry can predict patient outcome in urologic malignancies. Application of this analytical method in prostate cancer treated with radiation therapy is limited. We have evaluated the predictive ability of nuclear morphometry in such patients. Histologic sections from 23 men with clinically localized adenocarcinoma of the prostate treated with radiation therapy were studied. Nuclear morphometric parameters were assessed using a previously described and validated system. Univariate and multivariate logistic regression analyses and a Cox proportional hazards model were used to assess the ability of nuclear morphometric parameters to predict recurrence and disease-free interval. Ten patients had no recurrence with median follow-up of 47. 5 months, while 13 had recurrence. Gleason grade was not predictive of treatment outcome. Pre-treatment PSA data, available for only 11 patients, were predictive of treatment outcome. Several nuclear morphometric parameters predicted recurrence, including upper quartile of suboptimal circle fit and upper quartile of feret-diameter ratio. A prognostic factor score incorporating these 2 parameters was derived, which predicted disease-free interval (p = 0.0014). Int. J. Cancer (Pred. Oncol.) 84:594-597, 1999.

Clinical experience with an endobronchial implant

PURPOSE

To report clinical experience with an implantable capsule for treating endobronchial carcinoma by means of bronchoscopic insertion and retrieval.

MATERIALS AND METHODS

The capsule consists of a plastic cylinder containing high-activity iodine-125 seeds and four restraining legs. Twelve patients with recurrent (n = 11) or inoperable carcinoma (n = 1) received a median dose of 4,500 (range, 2,633-6,299) cGy at a 1-cm radius from the center of the implant.

RESULTS

No acute toxicities were observed. Five patients had complete regression of the endobronchial tumor, evidenced at bronchoscopy 2 months after therapy, and four patients had partial regression. Three patients failed to return for bronchoscopy. Symptomatic relief was achieved in eight of 10 patients with dyspnea, four of nine with cough, and two of three with hemoptysis. The median survival of the 12 patients was 6 months. The actuarial 1-year survival rate was 25%.

CONCLUSION

This capsule is safe and efficacious in treating recurrent or inoperable carcinoma of the lung.

Complications of radiation therapy: CT evaluation

Radiation therapy is an important technique for treating cancer. In the evaluation of the results of radiation therapy with computed tomography (CT), radiation-induced injuries to normal tissues are often detected. Common complications include pneumonitis, calcified lymph nodes, gastric ulceration, enteritis, hepatitis, cystitis, nephritis, osteitis, and insufficiency fractures. Rare complications include spontaneous pneumothorax, thymic cysts, vascular calcifications, and osseous sarcomas. Radiation-induced injury can usually be diagnosed from characteristic CT appearances and knowledge of the radiation port, radiation dose, and time interval since therapy. CT findings that cannot be explained on the basis of radiation therapy or that are suggestive of recurrent disease must be further evaluated.

Intraoperative I-125 seed implantation for extensive recurrent head and neck carcinomas

From 1978 to 1988, 41 patients with extensive recurrent carcinomas of the head and neck were treated with surgical resection plus intraoperative iodine-125 seed implantation. Surgery was performed to resect the tumors and to expose the tumor beds for implantation. I-125 seeds were implanted intraoperatively, with a spacing of 0.75-1 cm between adjacent seeds, either into the soft tissue in the tumor bed or onto small patches of gelatin sponges to cover the bone, nerve, or blood vessel involved with disease. Reconstructive flaps were used in 18 patients. The average I-125 dose delivered by the implanted seeds was 8,263 cGy. The determinate 5-year actuarial survival rate for the entire group was 40%. The 5-year local disease control rate was 44%. Major complications were transient wound infection (32%), flap necrosis (24%), fistula formation (10%), and carotid blowout (5%). These results indicate that surgical resection plus I-125 seed implantation provides a potentially curative treatment for patients with extensive recurrent head and neck carcinomas that would be considered traditionally unresectable and that would be treated only with palliative therapy.

Multiple spinal metastases from paraganglioma

Isolated vertebral body metastases from paraganglioma are exceedingly rare. They have been reported to occur in the presence of active primary tumor in the neck, local recurrence, or widespread metastases. A unique case of carotid body tumor (paraganglioma) is reported with the following features: (1) multiple vertebral body metastases (C6, T9, and L3) presenting with spinal cord compression, and no evidence of local recurrence or other metastatic disease; (2) absence of mitoses on the original specimen or the metastatic deposit; and (3) a prolonged interval (9 years) to the development of symptomatic metastases.

Pretreatment evaluation of sexual function in patients with adenocarcinoma of the prostate

Forty-three patients with adenocarcinoma of the prostate and available partners were interviewed to assess qualitatively and quantitatively their levels of sexual functioning prior to radiotherapy. The mean age was 67.7 years (58-80 years). The Derogatis Interview for Sexual Functioning (DISF) was the evaluation used. The Derogatis Interview for Sexual Functioning measures five domains of sexual functioning: sexual fantasy, arousal, experience, orgasm, and drive. Twenty-seven of the 43 (62.7%) patients evaluated were considered impotent. Fifteen of the 27 patients scored low in all five domains of sexual functioning even though they could achieve erection but they were unable to maintain erection throughout the phases of the sexual response cycle. Twelve of the 27 patients had scored 0 on sexual arousal and orgasm, thus had no ability for erection. Sixteen of the 43 (37.2%) patients were considered potent. This group of patients had achieved erection throughout the phases of sexual cycle and scored adequately in all five domains of sexual functioning. Patients with DISF score less than 20 were impotent. Those with a Derogatis Interview for Sexual Functioning score of greater than 45 were potent. Only 6 of 19 patients with scores between 20 and 45 were potent. The Derogatis Interview for Sexual Functioning score was highly prognostic for impotence, (p = .002) was easy to use and could be used for follow-up of the effect of therapy on sexual function in patients with adenocarcinoma of the prostate. Patients who present for radiation therapy are older, 50% are on cardiac or antihypertensive medication, and the majority (62.7%) are already impotent prior to therapy.

Pre and posttreatment evaluation of sexual function in patients with adenocarcinoma of the prostate

Twenty-seven patients with adenocarcinoma of the prostate, and available partners, were interviewed to qualitatively and quantitatively assess their level of sexual function prior to and 12 months after radiotherapy. Assessments were made using the Derogatis Interview for Sexual Functioning (DISF). Five domains of sexual functioning are measured: sexual fantasy, arousal, experience, orgasm, and drive. Prior to therapy 17 of 27 patients (62.9%) were considered impotent. There were eight patients with a DISF score of less than 20 who were impotent. Six patients had a DISF score of greater than 47 and were considered potent. Of the patients with DISF scores between 20-47 four were potent, and nine were impotent. Post radiation therapy three of the patients considered potent (with a score greater than 47) maintained their potent status. Four patients considered impotent prior to therapy became potent after therapy. All patients with a score less than 20 prior to radiation therapy remained impotent after therapy. Results indicate that an objective evaluation of sexual function pre treatment is necessary to determine the effect of radiotherapy. Our method of qualitative assessment of sexual function was easy to implement, was reproducible and could be used to evaluate long-term effects of radiotherapy on sexual function. Of the patients presenting for radiotherapy, 62.9% were impotent. Twelve months after radiation therapy 19 of 27 (70.3%) were impotent.

Staging, treatment, and results in testicular seminoma. A 12-year report

Sixty-one patients with histologically confirmed seminoma of the testis were treated by radiotherapy at The Johns Hopkins Hospital from 1975 to 1987. Fifty-seven patients remain disease-free, three patients died of intercurrent disease, and one patient (Stage IIB) died of widespread seminoma. Median follow-up for these patients is 5.5 years. Using a modified M.D. Anderson Hospital clinical staging system, 42 (69%) were Stage I, 16 (26%) were Stage IIA, and three (5%) were Stage IIB. Radiologic staging included both lymphangiogram (LAG) and abdominal computed tomography (CT) scan. Among Stage I disease, nine patients had false-positive CT scans, determined by negative LAG. Six patients had false-negative CT scans and were upstaged to Stage IIA by LAG. Treatment portals were altered in 15 of the 39 patients (38%) who had both LAG and CT scan. Overall actuarial survival (Kaplan-Meier method) was 97% at 5 years and 92% at 10 years. Five-year survival corrected for intercurrent disease was 100% for Stage I, 100% for Stage IIA, and two of three in Stage IIB patients. There were two distant treatment failures among the entire cohort. One patient who had Stage I disease was salvaged with local-field radiation and chemotherapy and is now without evidence of disease for 6 years. The second patient with Stage IIB seminoma receiving the same treatments disseminated and died. There were no significant acute toxicities or serious complications. In summary, proper staging with information gained from LAG and adequate radiation dose led to a 92% 10-year disease-free survival.

Internal mammary lymphadenopathy in breast carcinoma: CT appraisal of anatomic distribution

Internal mammary lymph nodes are an important site of occult metastasis in clinically operable and recurrent breast carcinoma. Anatomic distribution of enlarged internal mammary nodes in patients with breast cancer was analyzed in a review of thoracic computed tomographic studies of 219 women with operable, advanced, or recurrent cancer. Enlarged nodes were observed in 45 patients (20.5%), 32 with unilateral and 13 with bilateral involvement. Mean nodal diameter was 1.95 cm (0.6-6.0 cm). Lymphadenopathy was limited to one anterior intercostal space in 43%, two spaces in 26%, three spaces in 22%, and four spaces in 9%. Solitary or dominant nodal enlargement was centered at the first space in 14%, second space in 60%, and third space in 26%. Isolated lymphadenopathy in the fourth or fifth spaces was not observed. Metastases to internal mammary nodes frequently occur at multiple levels and are most common in the second and third spaces. This finding concurs with current surgical practice when internal mammary nodes are sampled.

The use of transmission block in the radiation therapy portal treatment of the inguinal nodes in late stage pelvic malignancies

Several pelvic malignancies in late stage have involvement of the inguinal nodes requiring radiation therapy. In recent years, our treatment plan has been changed to an opposed pelvic portal with the inguinal nodes treated in the anterior portal only by use of an anterior transmission block. This treatment technique was designed to avoid the problem of matching two radiation therapy portals. The advantages of this technique include easy treatment planning, adequate everyday set-up of the treatment portals, and no overlap by using separate pelvic fields and inguinal fields, reducing the complication rate.

Carcinoma of the prostate in the elderly: the therapeutic ratio of definitive radiotherapy

Withholding or reducing the intensity and aggressiveness of treatment for elderly cancer patients is a widespread tenet lacking substantiation in the literature. To assess the potential value of definitive external beam radiotherapy in the elderly, an analysis of the therapeutic ratio between local regional control and complications was performed in 34 prostatic cancer patients more than 75 years old. Median followup was 5 years (range 2 to 8 years). The 5-year actuarial local regional control rate was 91 per cent. The 5-year actuarial survival rate was 81 per cent and the 5-year survival rate free of disease was 63 per cent. There were no severe complications. Mild to moderate chronic complications occurred in 3 patients (9 per cent). This treatment resulted in an excellent therapeutic ratio, which demonstrates that external beam radiation can be given to elderly patients with acceptable morbidity and gratifying results.

The correlation of pretreatment transurethral resection of prostatic cancer with tumor dissemination and disease-free survival. A univariate and multivariate analysis

A multivariate Cox's hazard function analysis was performed on the prognostic variables selected from 240 patients with localized carcinoma of the prostate who received external beam radiotherapy to analyze the association between the method of biopsy and disease-free survival. The patients received 4500 cGy to the pelvis followed by a 2-week treatment rest and then an additional 2000 cGy to the reduced prostatic volume. Median follow-up was 4 years (range, 1-9 years). The 5-year actuarial local control rate was 91%. There was no difference in local control relative to the method of biopsy (needle biopsy or transurethral resection of the prostate [TURP]). The 5-year actuarial risks of metastases were 28% and 48% for the needle biopsy and TURP groups, respectively (P less than 0.01). The 5-year disease-free survival for the needle biopsy and TURP groups were 55% and 42%, respectively (P less than 0.05). This difference maintained statistical significance for Stage C and Grade III tumors. A multivariate analysis demonstrated that "method of biopsy" was the third most powerful variable after serum acid phosphatase level and modified Broders' grade in predicting disease-free survival. Patients who had TURP had an almost twofold higher relative risk of disease progression than those who had needle biopsy. This study established the correlation of the method of biopsy with a lower probability of disease-free survival over and above the information obtained from the clinical stage, histologic grade, presence or absence of symptoms of obstruction, acid phosphatase level, and lymph node status. The association of TURP with tumor dissemination is discussed with reference to the mechanism of metastasis formation and prevention.

Computed tomography in radiotherapy planning of the axillary region

Computed tomography was used in the pretreatment evaluation of three patients with neoplasms involving the axillary region. Radiation therapy treatment fields were more accurately defined using the computed tomography data. Computed tomography has a valuable role in radiotherapy planning in patients with tumors involving the axillary region.

CT of the axilla: normal anatomy and pathology

CT of the axilla is useful in evaluating a palpable mass or brachial plexopathy, staging a known malignant neoplasm and evaluating a radiation therapy port.

Definitive radiotherapy following prostatectomy: results and complications

Thirty-four patients with carcinoma of the prostate treated by prostatectomy received postoperative external beam radiation. Sixteen patients were treated within 4 months of radical prostatectomy (group 1), 12 patients were treated for prostate carcinoma following initial enucleative prostatectomy for benign hypertrophy (group 2) and 6 patients were treated for palpable local recurrence 4 to 10 years following radical prostatectomy (group 3). The indications for postoperative radiotherapy following radical prostatectomy included extracapsular extension, seminal vesicle invasion, peri-prostatic soft tissue involvement, positive margins or palpable local recurrence. Eighty-five percent of the patients received whole pelvic radiation. All patients then had a 2-week treatment rest followed by a reduced portal to the prostate bed to a dose of 6500 cGy. The local control rate after radiotherapy was 100% with a median follow-up of 4 years. The 5-year actuarial survival and disease-free survival rates for all patients were 82 and 72%, respectively. In group 1, the 5-year actuarial survival and disease-free survival rates were 100 and 91%, respectively. In group 2, these rates were 77 and 64%. Three of the six patients in group 3 died within 30 months of radiotherapy. Fourteen patients (41%) had mild to moderate treatment related symptoms including seven patients (21%) with lower extremity or genital edema, five patients (15%) with urinary stress incontinence, two patients (6%) with urethral stricture and three patients (9%) with proctitis. Six of eight patients who were potent prior to radiation retained potency thereafter. No severe complications occurred. We conclude that external beam radiation therapy administered after prostatectomy resulted in an acceptable therapeutic ratio with 100% local regional control, and an acceptable complication rate (41%).

Computed tomography in radiation therapy treatment planning of hepatic metastases

Forty-four patients with symptomatic hepatic metastases from a colon primary had CT prior to radiation therapy treatment planning. A control group of 44 patients with no evidence of liver metastases was analyzed for comparison. The objectives of planning were to deliver as homogeneous a dose to the whole liver as possible and no treat one kidney or more than one-half of both kidneys. Conventional anteroposterior/posteroanterior portals were found to be inadequate for the treatment of 60% of patients with metastases and 10% of patients with no liver involvement. Among the metastatic group, 50% required oblique planning, 40% anteroposterior/posteroanterior and 9% posteroanterior and left lateral portals to meet the treatment planning objectives. Among the control group of patients with no liver metastases, only 9% required oblique portals and 8% could not be treated because of left hydronephrosis or a solitary right kidney. It is concluded that all patients receiving radiation therapy to the liver for symptoms or prophylaxis require CT for optimum radiation therapy treatment planning.

New frontiers in the treatment of lung cancer

Surgical resection still is the only significant curative approach in nonsmall cell lung cancer. Recent surgical experience indicates that a modest decrease in the death rate from bronchogenic carcinoma may occur in three general areas: (1) the detection and treatment of radiographically occult squamous cell carcinoma; (2) the combination of adjuvant chemotherapy and surgical excision in selected patients with small cell carcinoma; and (3) surgical resection and postop irradiation of patients with hilar and mediastinal lymph node metastases. At the time of diagnosis, 80 to 85% of the patients present with unresectable lung cancer. These patients may benefit from other modalities of therapy, i.e., radiotherapy, chemotherapy, or immunotherapy. Failures following radiotherapy in unresectable nonsmall cell lung cancer are due to (1) distant metastasis, (2) local region failure, and (3) local and distant failure. To increase the local control, new methods of treatment have been tried, such as hyperfractionation of radiotherapy and the use of 131I antiferritin immunoglobulin. The development of effective systemic chemotherapy is necessary to treat metastatic bronchogenic carcinoma. The response rate to chemotherapeutic agents is substantially lower in nonsmall cell carcinoma than in small cell carcinoma. Investigation is ongoing to assess the effectiveness of new antitumor drugs used alone, in combination with other drugs, or combined with other modalities for the treatment of bronchogenic carcinoma.

Computed tomographic diagnosis of radiation ileitis

Two patients were studied by CT after resection of colon carcinoma followed by abdominal radiation therapy. Based upon the findings of a localized, mass-like confluence of shortened loops of bowel with thickening of bowel wall, adherence of adjoining loops, thickening of the adjacent mesentery, and increased density of mesenteric fat the patients were considered to have radiation ileitis. Recurrent tumor was excluded by autopsy and confirmed by a 1-year asymptomatic follow-up period.