An efficient attention encoder decoder-based residual-UNet for the segmentation of liver and lung tumour

Genetic Alterations in Cancer Knowledge System: Analysis of Gene Mutations in Mouse and Human Liver and Lung Tumors

The liver is an essential metabolic organ of the human body, and malignant liver tumors seriously affect and threaten human life. The segmentation algorithm for liver and liver tumors is one of the essential branches of computer-aided diagnosis. This paper proposed a two-stage liver and tumor segmentation algorithm based on the convolutional neural network (CNN). In the present study, we used two stages to segment the liver and tumors: liver localization and tumor segmentation. In the liver localization stage, the network segments the liver region, adopts the encoding–decoding structure and long-distance feature fusion operation, and utilizes the shallow features’ spatial information to improve liver identification. In the tumor segmentation stage, based on the liver segmentation results of the first two steps, a CNN model was designed to accurately identify the liver tumors by using the 2D image features and 3D spatial features of the CT image slices. At the same time, we use the attention mechanism to improve the segmentation performance of small liver tumors. The proposed algorithm was tested on the public data set Liver Tumor Segmentation Challenge (LiTS). The Dice coefficient of liver segmentation was 0.967, and the Dice coefficient of tumor segmentation was 0.725. The proposed algorithm can accurately segment the liver and liver tumors in CT images. Compared with other state-of-the-art algorithms, the segmentation results of the proposed algorithm rank the highest in the Dice coefficient.

The chlorinated hydrocarbons chloroform (CHCl3), 1,1-dichlorethane (1,1-DCE) and 1,2-dichloroethane (1,2-DCE) have been detected in finished drinking water. When administered to B6C3F1 mice by gavage in corn oil, these compounds have been shown to induce hepatic tumors. The present study examines the effect on liver tumor incidence of continuous treatment of CHCl3 (600 mg/L and 1800 mg/L), 1,1-DCE (835 mg/L and 2500 mg/L), and 1,2-DCE (835 mg/L and 2500 mg/L) administered in drinking water to male B6C3F1 mice using a two-stage (initiation/promotion) treatment protocol. Seventy 4-week-old male B6C3F1 mice constituted each treatment group. Of these mice, 35 were initiated by treatment with diethylnitrosamine (DENA) (10 mg/L) in the drinking water for 4 weeks. The remaining 35 received deionized drinking water. Each group was subsequently treated with one of two concentrations of CHCl3, 1,1-DCE, or 1,2-DCE in drinking water for 52 weeks. An additional group received phenobarbital (PB) (500 mg/L) and served as the positive control for liver tumor promotion. Mice were sampled after 24 weeks (10 mice) and 52 weeks (25 mice). At sampling, liver and lung tumors were detected. None of the compounds increased the number or incidence of lung or liver tumors by themselves. PB promoted liver tumor formation (but not lung tumors) in the DENA-initiated mice. 1,1-DCE and 1,2-DCE did not affect the incidence or number of liver or lung tumors in the DENA-initiated animals. CHCl3, however, inhibited liver and lung tumorigenesis in the DENA-initiated mice.

The chlorinated hydrocarbons chloroform (CHCl3), 1,1-dichlorethane (1,1-DCE) and 1,2-dichloroethane (1,2-DCE) have been detected in finished drinking water. When administered to B6C3F1 mice by gavage in corn oil, these compounds have been shown to induce hepatic tumors. The present study examines the effect on liver tumor incidence of continuous treatment of CHCl3 (600 mg/L and 1800 mg/L), 1,1-DCE (835 mg/L and 2500 mg/L), and 1,2-DCE (835 mg/L and 2500 mg/L) administered in drinking water to male B6C3F1 mice using a two-stage (initiation/promotion) treatment protocol. Seventy 4-week-old male B6C3F1 mice constituted each treatment group. Of these mice, 35 were initiated by treatment with diethylnitrosamine (DENA) (10 mg/L) in the drinking water for 4 weeks. The remaining 35 received deionized drinking water. Each group was subsequently treated with one of two concentrations of CHCl3, 1,1-DCE, or 1,2-DCE in drinking water for 52 weeks. An additional group received phenobarbital (PB) (500 mg/L) and served as the positive control for liver tumor promotion. Mice were sampled after 24 weeks (10 mice) and 52 weeks (25 mice). At sampling, liver and lung tumors were detected. None of the compounds increased the number or incidence of lung or liver tumors by themselves. PB promoted liver tumor formation (but not lung tumors) in the DENA-initiated mice. 1,1-DCE and 1,2-DCE did not affect the incidence or number of liver or lung tumors in the DENA-initiated animals. CHCl3, however, inhibited liver and lung tumorigenesis in the DENA-initiated mice.

Alterations at the Ink4a locus in transplacentally induced murine lung tumors

Liver Segmentation on a Variety of Computed Tomography (CT) Images Based on Convolutional Neural Networks Combined with Connected Components

2-Amino-3-methylimidazo[4,5-f]quinoline (IQ) is a very potent mutagen that is carcinogenic in rodents and nonhuman primates. IQ-induced CDF1 mouse lung and liver tumors were examined for activated Ki-ras and Ha-ras genes, respectively. Polymerase chain reaction (PCR)-amplified target DNAs were analyzed for mutations of codons 12, 13, and 61 by single-strand conformation polymorphism (SSCP) and direct sequencing methods. All mutations were localized to codon 61 of the ras genes. Forty-nine of 54 lung tumors induced by IQ possessed activating Ki-ras mutations, as did 20 of 26 lung tumors from the vehicle-treated animals; 80% and 75% of these mutations, respectively, were A-->T transversions of the second nucleotide redundant. One lung adenoma from the IQ-treated group contained a tandem duplication of the sequence corresponding to codons 50-57 of the Ki-ras gene (unpublished observations). In addition, seven of 34 IQ-induced liver tumors harbored activating Ha-ras mutations: five were C-->A (G-->T) transversions at the first nucleotide, and two were A-->T transversions at the second nucleotide of codon 61. None of the 15 liver tumors collected from the vehicle-treated mice possessed Ha-ras mutations in codon 12, 13, or 61. These data indicate that IQ induces Ha-ras gene activation in CDF1 mouse liver tumors. The mechanisms of lung tumor induction by IQ, however, is obscured by the high frequency of Ki-ras A-->T mutations observed in both the IQ-induced and spontaneous lung tumors. The different ras mutational spectra in lung and liver tumors may suggest either that two different pathways of IQ metabolism exist in these organs or that IQ contributes to CDF1 lung tumorigenesis by a mechanism other than its direct interaction with the Ki-ras gene.

Fluoranthene (FA), a major environmental pollutant, induced lung and liver tumors 6-9 months after intraperitoneal injection of 0.7, 1.75 and 3.5 mg FA into preweanling CD-1 mice. There was a dose-dependent increase in lung tumors with a maximum tumor incidence of nearly 45% and a maximum tumor multiplicity of 0.6-0.7 lung tumors/mouse. No significant difference was noted in lung tumors in the 6 and 9 month bioassays, although fewer tumors were consistently noted in mice treated with the two lowest doses of FA. Indices of lung tumor incidence (ED50) and multiplicity (TM1.0) were similar for the two bioassays and ranged from 18.9-19.5 and 26.2-27.2 mumol respectively. Male mice had up to two times more lung tumors than females but these results were not statistically significant. Liver tumors (nodular hyperplasia) appeared only in FA-treated males but no dose-response relationship was evident. However, liver tumors were observed in only 0-10% of the male mice in the 6 month treatment groups, but in 20-60% of the males in the 9 month groups. Because the CD-1 preweanling mouse responded to the weak lung tumorigen FA, it is a viable, limited-term bioassay for measuring tumorigenicity of PAH and combustion emissions.

The effect of phenobarbital on simultaneous induction of liver and lung tumors was examined in inbred DDD mice. Group 1 of newborn mice received a single intraperitoneal (ip) injection of dimethylnitrosamine (DMN) and after weaning they were given 0.05% phenobarbital solution to drink. Group 2 received an injection of DMN like Group 1 but were then given normal water. Group 3 were injected ip with 0.9% NaCl solution and then given phenobarbital solution to drink as in Group 1, and Group 4 were injected with 0.9% NaCl solution like Group 3, and then given tap water to drink. The animals were examined 16 weeks after birth. In Group 1, 27 of 35 mice (77%) had liver tumor and 15 (43%) had lung tumor. In Group 2, 8 of 24 mice (33%) had liver tumor and 16 (67%) had lung tumor. Animals in Groups 3 and 4 did not develop tumors. The difference in the incidences of liver tumor, but not lung tumor, in Groups 1 and 2 was statistically significant (P less than 0.01); that is, a promoting effect of phenobarbital was observed in induction of liver tumor, but not lung tumor.

Transplacental carcinogenesis represents a good model in which to study the involvement of tissue-specific oncogene activation in carcinogenesis because a single exposure to a carcinogen induces tumors at various sites. We tested tumors of the skin, liver, and lung produced in mice after transplacental 7,12-dimethylbenz[a]-anthracene (DMBA) exposure for possible activation of ras genes. XbaI restriction fragment-length polymorphism analysis has shown that exposure to DMBA in utero may result in appearance of A----T transversion at the second position of codon 61 of Ha-ras oncogene in skin and liver tumors but not in lung tumors. Moreover, DNA samples isolated from spontaneous and DMBA-induced lung and liver tumors were analyzed for mutations at the same position of Ki-ras oncogene using differential hybridization with specific oligonucleotides. Among five spontaneous lung tumors, three cases of A----G transition, and one case of A----T transversion were found, whereas four of ten lung tumors of DMBA-treated animals were positive for A----T mutation. No Ki-ras mutation was detected in one spontaneous and four DMBA-induced hepatomas. In two cases, we revealed Ki-ras A----T mutation in the lung tumor and Ha-ras mutation in the liver tumor taken from the same animal. These results indicate first that DMBA treatment may induce A----T mutation at the second position of codon 61 both in Ha-ras and in Ki-ras and, second, that the role of different activated oncogenes in carcinogenesis may differ, depending on the tissue in which the tumor develops.

Transplacental Initiation of Liver, Lung, Neurogenic, and Connective Tissue Tumors by <i>N</i>-Nitroso Compounds in Mice

Benz(a)anthracene injected subcutaneously during the first 3 days of life caused a dose related increase in the incidence of liver and lung tumours in Swiss mice but over a similar dose range, the K region epoxide of benz(a)anthracene was less effective. Neonatally injected 7-methylbenz(a) was considerably more active than its K region epoxide in increasing the incidence of liver tumours in males. Both the parent compound and the epoxide slightly raised the incidence of lung tumours. Both chrysene and its K region epoxide increased liver tumour incidence but not lung tumour incidence. The K region epoxides of dibenz(a,h)-anthracene and 3-methylcholanthrene were without apparent effect on the incidence of liver, lung or other tumours despite indications from previously reported studies that the parent hydrocarbons are active at the same dose levels. The K region epoxide of phenanthrene had no effect on the incidence of any kind of neoplasm.

Polychlorinated biphenyls (PCB), which are tumor promoters, have been found in human tissues for decades. Their contribution to cancer risk may only now start to appear, due to long human cancer latency and the nature of tumor promotion. Epidemiological associations have been seen between PCB exposure or tissue content and cancer at several sites. In rodents, tumor promotion by PCBs has been little studied in tissues other than liver. Previously, in an experiment modeling infant carcinogen exposure following PCBs received in milk, lung and liver tumors, initiated neonatally in mice by the environmental nitrosamine N-nitrosodimethylamine (NDMA), were promoted by later treatment with Aroclor 1254. The present study was undertaken to confirm and characterize the effects of Aroclor 1254 on tumor number, latency, size and malignancy. Male Swiss mice were given NDMA on postnatal day 4 and Aroclor 1254 (250 mg/kg) on day 8, and killed at intervals. Eight PCB congeners were quantified in the carcasses. Incidences of mice with NDMA-initiated lung tumors at 28 weeks of age were increased 2.5-fold by PCBs. Multiplicities of lung tumors were enhanced four-fold by PCBs at 28 and 52 weeks. By 72 weeks tumor numbers were similar in the NDMA-only and NDMA-PCB groups. Liver tumors first occurred in significant numbers at 52 weeks and only in mice receiving both NDMA and PCBs. As for the lung, at 72 weeks the incidence was high in both the NDMA-only and NDMA-PCB groups. Sizes of tumors and liver carcinoma incidence were not altered by PCB treatment. Carcass analysis revealed a significant positive association between lung tumor numbers at 28 weeks and relative percentage of 2,2',4,4',5-pentachlorobiphenyl, with no other correlations. The results confirm that PCBs promote lung as well as liver tumors, by triggering the early appearance of latent initiated tumors otherwise presenting in old age.

No. 163: Comparison of Two Different Methods for Inoculating VX2 Tumors in Rabbit Livers and Hind Limbs

Several nitrosamines and an azoxyalkane have been administered intravesically to groups of 12 female F344 rats, twice a week for 20 or 30 weeks. Many of the nitrosamines were as efficacious in giving rise to the same tumors of internal organs as when similar doses were administered orally, showing that absorption from the bladder was as rapid as from other sites. The tumors produced included lung and kidney tumors by nitrosodimethylamine, colon and Zymbal gland tumors by azoxymethane, liver tumors by methylnitrosoethylamine (but not by nitrosodimethylamine), liver and esophagus tumors by nitrosodiethylamine, liver and lung tumors by methylnitrosamino‐3‐pyridylbutanone, liver tumors by nitrosomorpholine, and tumors of the esophagus by methylnitroso‐n‐butylamine, 2,6‐dimethylnitrosomorpholine and methylnitrosamino‐N,N‐dimethylethylamine. Bladder tumors were induced by intravesicular administration of only low doses of nitrosobis‐(2‐oxopropyl)amine and to a lesser extent by methylnitroso‐n‐hexylamine and nitroso‐(2‐hydroxypropyl)(2‐oxopropyl)amine, which all induced tumors systemically in addition. The bladder mucosa seemed to lack enzymes necessary to activate most nitrosamines to locally acting proximate carcinogens, but was quite transparent to the passage of carcinogenic nitrosamines present in the urine into the body to induce tumors in distant organs.